U.S. Department of Health and Human Services

Cash and Counseling: Consumer's Early Experiences in Arkansas

Leslie Foster, Randall Brown, Barbara Carlson, Barbara Phillips and Jennifer Schore

Mathematica Policy Research, Inc.

October 2000

This report was prepared under contract #HHS-100-95-0046 between the U.S. Department of Health and Human Services (HHS), Office of Disability, Aging and Long-Term Care Policy (DALTCP) and the University of Maryland. Additional funding was provided by the Robert Wood Johnson Foundation. For additional information, you may visit the DALTCP home page at http://aspe.hhs.gov/daltcp/home.htm or contact the ASPE Project Officer, Pamela Doty, at HHS/ASPE/DALTCP, Room 424E, H.H. Humphrey Building, 200 Independence Avenue, SW, Washington, DC 20201. Her e-mail address is: pdoty@osaspe.dhhs.gov.

This paper describes the experiences of 194 early clients in the Arkansas Cash and Counseling Demonstration, IndependentChoices. The description is based on the clients’ responses to a telephone survey conducted about nine months after they applied to enter the program and were randomly assigned to the demonstration’s treatment group to receive a monthly cash allowance. Mathematica Policy Research, Inc. (MPR), the demonstration evaluator, administered the survey.

After briefly describing the Arkansas project, data and methods used, and client characteristics, this paper discusses client outcomes in four areas: (1) program participation; (2) uses of services, goods, and cash; (3) hiring of caregivers and revision of cash expenditure plans; and (4) satisfaction. Particularly interesting results include the following:

• On average, recipients of the experimental cash benefit were elderly (73 percent were aged 65 or older), in poor health, and had high levels of functional disability. This finding contradicts the conventional wisdom that the option to purchase one’s own services out of an allowance in lieu of receiving professionally managed personal attendant services appeals primarily to younger adults with physical disabilities or to the elderly with less severe disabilities.

• Two-thirds of the cash benefit recipients were still participating in IndependentChoices after nine months. Nine percent had died prior to the nine-month interview and 24 percent had disenrolled.

• Disenrollments occurred for a variety of reasons, including factors that made some individuals ineligible to continue (e.g., loss of financial eligibility for Medicaid, moving out of state, entry into a nursing facility). Fifty-nine percent reflected a decision to return to the traditional system of receiving personal care services through agencies.

• Choosing to return to traditional services was more likely to occur during the first four months of participation; indeed, most such choices occurred after random assignment to the treatment group but before the participant began receiving the allowance.

• Very few participants elected to receive the full amount of their monthly cash allowance in “cash.” Nearly all those who employed individual attendants preferred to have the fiscal intermediary/counseling agency act as their payroll agent and file the employer’s share of applicable taxes. Described to participants as a “book-keeping” service, this option was offered to participants at no extra charge. Most participants chose to have the service maintain an account for them and make major purchases, in addition to paying workers, on their behalf.

• Most participants (92 percent) had at least one paid caregiver. Eighty-six percent of participants who received the cash allowance (that is, excluding those who died or disenrolled before actually receiving the allowance) used some or all of the allowance to hire caregivers.

• Over 90 percent of participants hired paid caregivers who were family members, friends, or neighbors. Only a handful of participants hired individuals previously unknown to them.

• Most participants had live-in help (paid or unpaid) and most also had both paid and unpaid caregivers. However, substantial minorities of participants lived alone (over one-third) or did not have any unpaid helpers (one-quarter).

• Prior to spending their cash allowances, participants were required to develop cash expenditure plans. Actual spending was required to follow the plan, in order to avoid possible overspending or unauthorized purchases; however, participants could revise their plans at any time as long as they did not go over-budget. Forty percent of participants who accessed the allowance revised their cash expenditure plans at least once prior to the nine-month interview.

• Eighty-two percent of participants who received the cash allowance reported that it improved their quality of life. None said that they were worse off.

• Roughly half those reporting an improved quality of life cited reasons having to do with caregivers (e.g., being able to hire individuals of their choice, having better access to the right kind of personal assistance services, or more conveniently scheduled services, and being able to relieve family members). Almost half cited the flexibility to use the allowance to purchase other kinds of goods and services (e.g., medications, medical and personal care supplies, equipment, home modifications).

• Participants were allowed to “save up” from month to month in order to accumulate enough funds to purchase a major piece of equipment or home modification. Relatively few clients (10 percent) used the allowance to purchase or repair equipment for personal activities other than meal preparation or housekeeping. Similarly, only 10 percent of participants used the allowance to make a home modification and only 2 percent used the allowance to make a vehicle modification.

• All respondents expressed satisfaction with their relationships with paid caregivers who had helped them recently.

• More than 9 out of 10 participants (including those who disenrolled and family members responding on behalf of those who died) would recommend IndependentChoices to others seeking greater control over their personal care services.

A. CASH AND COUNSELING IN ARKANSAS

IndependentChoices gives beneficiaries who are eligible for Medicaid personal assistance services (PAS) a chance to receive a monthly cash allowance in lieu of traditional services.¹ Beneficiaries may use the allowance to hire caregivers or purchase equipment that would enhance their ability to live independently. As part of the demonstration’s evaluation, eligible beneficiaries interested in receiving the allowance are randomly assigned by the evaluator either to the treatment group (which receives the cash allowance) or to the control group (which continues to receive traditional Medicaid PAS). This paper examines only the responses of beneficiaries assigned to receive the allowance. We refer to them as “clients” throughout the paper.²

In addition to the monthly allowance, the demonstration provides counseling services (for example, to help the client develop an allowance expenditure plan) and bookkeeping services (for example, to pay and withhold taxes for caregivers hired with the allowance). Clients may receive these services from one of two agencies in Arkansas, depending on where they reside. Clients who are unable to manage their own PAS (for example, to make decisions about whom to hire and how much to pay) may have a representative do it for them. A representative may be a family member, friend, legal guardian, or other legally appointed individual.

B. DATA AND ANALYTIC APPROACH

The data for this analysis were drawn from computer-assisted telephone interviews with demonstration clients, including disenrollees. This paper summarizes the responses of 194 individuals who enrolled in the Arkansas demonstration between its inception (December 1998) and June 1999, were randomly assigned to receive the cash allowance, and completed a nine-month follow-up interview between September 1999 and March 2000. The nine-month interview consists of questions with precoded answers, with a few exceptions: an additional set of questions about disenrolling is asked of clients who dropped out of the program between the four- and nine-month interviews.

All but 20 sample members had also completed an interview about four months after random assignment. That interview asked treatment group members about their early program experiences, such as developing cash expenditure plans, hiring caregivers, and reasons for disenrolling. To preserve comparability between the treatment and control groups, we conducted nine-month interviews even with disenrolled clients and with the proxies of deceased clients.

Questions in the nine-month interview refer to a variety of reference periods, including the present, the most recent two weeks the client was in the home (as opposed to in a hospital, nursing home, or long-term care facility), the entire nine months since enrollment into the demonstration, and the period between the four- and nine-month interviews. For example, we asked about clients’ current unmet needs and satisfaction with care, as reports of past perceptions are unlikely to be accurate. Questions about difficulty with daily activities and help performing daily activities refer to a recent two-week period. We set a short recall period because clients performed daily activities frequently, and we sought detailed information about the care received. In contrast, we asked clients about community services or equipment purchases during the entire nine-month period because these events were likely to be relatively infrequent. Questions about revising cash expenditure plans refer to the period between the four- and nine-month interviews.

The tables in this paper present percentage distributions, frequencies, and cross-tabulations of selected survey responses. Because our goal is simply to describe clients’ experiences, we do not present standard errors, confidence intervals, or tests of hypotheses, nor do we compare clients with the control group.³ Many of the tables categorize responses by clients’ age groups and by how long clients had been using PAS at the time they enrolled in the study. In the tables, clients are divided into two age groups: (1) age 65 or older, and (2) younger than age 65. Clients who received PAS for three or more months prior to enrollment are distinguished from “new” users--those who had received PAS for fewer than three months. Given the small sample available for this analysis, we note only differences between subgroups that are 15 percentage points or larger. Smaller differences may be due to chance.

Nine-month interviews were completed by 202 eligible sample members, yielding an interim response rate of 90.2 percent (see Table 1).⁴ (Eight of the completed interviews required telephone call-backs to collect missing data and were dropped from this analysis, leaving 194 cases. The eight cases will be included in a future analysis.) About half the 194 interviews used in this analysis were completed by the clients themselves; the other half were completed by proxy respondents (not shown). Thirty-one percent of the proxies were also paid caregivers.

C. CLIENT CHARACTERISTICS AND UNPAID HELP

On average, clients were elderly, in poor health, and had high levels of functional impairment. Seventy-three percent of the respondents in the cohort were age 65 or older, and 77 percent had been using PAS for at least three months before enrolling in IndependentChoices (see Table 1). A majority (54 percent) rated their health as poor relative to the health of other people their age, and 84 percent had at least one chronic health condition (see Table 2). Without help, bathing would have been very difficult or impossible for 70 percent of clients. Getting in or out of bed or using the toilet would have been very difficult or impossible for about one-third.

Substantial fractions of clients lived alone or did not receive unpaid help. Overall, more than one-third lived alone, and one-quarter had no unpaid caregivers (see Table 3). The three-quarters who had unpaid help were split about evenly into groups defined by whether they received unpaid help from one, two, or three or more people. Half the clients who reported receiving unpaid help from a relative received this help from daughters or daughters-in-law (not shown). Most (59 percent) of the clients who had at least one unpaid caregiver did not live with the caregiver. New PAS users were especially likely not to have live-in unpaid caregivers (74 percent did not).

D. PROGRAM PARTICIPATION

Two-thirds of the clients in this cohort were still participating in IndependentChoices after nine months. Disenrollment rates were substantially higher for older clients than for younger ones (40 percent versus 14 percent) and for prior PAS users than for new users (37 percent versus 21 percent). About 80 percent of all clients, including disenrollees, had received the monthly allowance by the time of the nine-month interview (see Table 4). Approximately 98 percent of those still enrolled at that time had received the monthly allowance, and only about 40 percent of those who left the program had begun to receive the allowance before leaving (see Table 5). By the nine-month interview, one-third of the analysis sample (64 clients) was either deceased (18 clients) or had disenrolled from IndependentChoices (46 clients) (not shown). Of these, most (44 clients) had died or disenrolled before the four-month interview.

E. USES OF SERVICES, GOODS, AND CASH

1. Personal Assistance Services

Most clients in the sample received personal assistance from a paid caregiver. About 92 percent of the 172 clients who were home for at least two weeks shortly before the interview received help during that period from a paid caregiver, including agency workers (see Table 6). Of the clients with paid caregivers, 60 percent had one paid caregiver, 27 percent had two, and 13 percent had three or more. Younger clients were more likely to have just one paid caregiver (86 percent of these clients). Twenty-six percent of clients with paid caregivers had live-in paid caregivers, and 87 percent had visiting paid caregivers. Although Arkansas Cash and Counseling clients cannot hire a spouse or legal guardian as a caregiver, 59 percent of clients with paid caregivers did receive paid help from another relative. New PAS users were somewhat more likely to have hired a relative (79 percent did so). Of the 93 respondents with a paid caregiver who was a relative, 59 percent identified this caregiver as a daughter or daughter-in-law (not shown).

The highest percentage of clients with visiting paid caregivers (40 percent) received between 21 and 40 hours of care from all paid sources during the two weeks about which they were asked (see Table 7). About one-third received between 10 and 20 hours of care from these sources, and 15 percent received more than 40 hours of paid care during the two weeks. Similarly, of the 40 clients with live-in paid caregivers, 16 clients (40 percent) received between 21 and 40 hours of care from that person (not shown).

Paid caregivers helped with all aspects of their clients’ care, though less often with routine medical care than with personal care or household and community activities (see Table 8). Ninety-five percent of all clients paid to have someone help with light housework, and 90 percent paid for help bathing. Eighty percent paid for help preparing meals, 62 percent relied on paid caregivers for help getting in or out of bed, and more than half relied on them for help taking medicine or with other routine health care, such as checking blood pressure. Younger clients were more likely to receive paid help with transportation than were older clients (67 percent versus 36 percent). New PAS users reported similarly high rates of paid help with transportation but reported less paid help with several types of personal care than did clients with more PAS experience (for example, 48 percent of new PAS users received help getting out of bed, compared with 65 percent of those with longer PAS tenures).

2. Goods and Community Services

Since enrollment, about one-third of clients had obtained or repaired equipment for personal activities, communication, or safety, and about 10 percent reported using their monthly allowance to pay for this equipment (see Table 9). Fewer clients obtained or repaired equipment to prepare meals or keep house (22 percent), and 7 percent paid for their equipment with the monthly allowance. One-quarter of all clients had modified their homes since enrolling in the program, and 10 percent reported paying for the modification with the monthly allowance. Clients who had been receiving PAS for three months or longer prior to enrollment were nearly twice as likely as those with shorter tenures to have modified their homes (see Table 10). Only a few clients (4 percent) had modified their vehicles to make them easier to ride in.

Since enrolling in IndependentChoices, clients used some community services substantially more than others. About 36 percent of clients had received home-delivered meals since enrollment; a substantially larger percentage of older clients (46 percent) than younger clients (6 percent) received these meals (see Table 11). Seventeen percent used special transportation services to see a physician or perform errands. In contrast, fewer than 10 percent had attended social or recreational programs since enrolling, and very few (2 percent) attended adult day care.

F. HIRING CAREGIVERS AND REVISING EXPENDITURE PLANS

1. Hiring Caregivers

Eighty-six percent of currently enrolled cash recipients used the allowance to hire at least one caregiver directly, rather than through an agency (see Table 12). Of those, 70 percent (85 clients) had hired their caregivers by the time of the four-month interview (data not shown).

This analysis sample includes 145 clients who tried to hire caregivers by the time of the nine-month interview (see Table 13). Of those, 75 percent attempted to hire a family member; 37 percent tried to hire a friend, neighbor, or church member; and 17 percent tried to hire a former home care agency worker. About one-fifth asked their families or friends to recommend a worker.⁵ Family members were the most commonly hired caregivers by far, followed by friends and neighbors (see Table 14). Very small percentages of clients hired caregivers who were not immediate acquaintances.

2. Revising Expenditure Plans and Other Activities Between Interviews

Forty percent of cash recipients had revised their cash expenditure plans between the time of the four- and nine-month interviews (see Table 15). Nearly one-quarter of clients who had hired someone since enrollment or who had tried to hire between interviews reported changing the wages, hours, or type of work with which they wanted help. Only 5 percent of clients who had hired a caregiver since enrollment and who received the monthly allowance provided any fringe benefits. None of the few clients providing fringe benefits at the four-month interview were providing additional benefits by the nine-month interview. Fewer than 2 percent had received peer counseling between interviews (not shown). (The same percentage reported having received peer counseling at the time of the four-month interview.)

G. SATISFACTION WITH LIFE AND QUALITY OF CARE

Clients were largely satisfied with their lives and care. Nearly all clients (96 percent), including disenrollees, expressed satisfaction with their overall care (see Table 16). Furthermore, nearly all clients (99 percent) were pleased with the way their paid caregivers fulfilled their duties, such as providing personal and routine health care. Importantly, every one of 129 respondents expressed satisfaction with the relationship he or she had with the paid caregivers who had helped them recently. Moreover, for each of these measures, large majorities of satisfied clients (from 79 percent to 94 percent) said they were “very satisfied” with that aspect of their care (data not shown). Eighty percent said their paid caregivers always or almost always completed their tasks. Although 95 percent were satisfied with the time of day they received help, about half said that it would be difficult to change their caregiver’s schedule if it were necessary to do so.

Despite the high levels of satisfaction reported, when asked about specific activities, substantial proportions of clients said they were not receiving as much help as they needed. Approximately 40 percent reportedly needed more help with meal preparation and housework (see Table 17). One-third had unmet transportation needs, and a similar percentage had unmet personal care needs. About one-quarter needed more help with medications and other routine health care. For each of these measures, however, clients who had received the cash allowance were less likely than those who had not received the allowance to report an unmet need. One-quarter of cash recipients needed more help with medication and routine medical care versus one-third of nonrecipients. Cash recipients were also less likely than nonrecipients to need more help with meals and housework (38 percent of recipients versus 51 percent of nonrecipients), transportation, and personal care (29 percent of recipients versus 43 percent of nonrecipients, for both measures). (Data not shown.)

About 80 percent of all clients said they were satisfied with their lives (see Table 17). Again, cash recipients were more likely than nonrecipients to be satisfied (82 percent versus 71 percent). Nonetheless, almost 90 percent of all clients reported that health problems or lack of assistance limited their recreational, cultural, social, or religious activities. Clients between the ages of 18 and 75 were asked whether their ability to pursue education or paid work was hindered by health or lack of assistance. Ninety percent of respondents could not fulfill educational pursuits, and 98 percent had limited ability to work for pay.

H. SATISFACTION WITH INDEPENDENTCHOICES

Clients found much to commend about IndependentChoices. More than 90 percent of them (including 100 percent of enrollees and 68 percent of disenrollees), said they would recommend the program to someone wanting more control over PAS. Overall, 82 percent said the monthly allowance improved their lives, and more than 75 percent reported a great deal of improvement (see Table 18). Not one respondent said that receiving the cash allowance reduced the quality of their lives.

Clients who said the monthly allowance improved their lives were asked (in an open-ended question) to name the most important way that it did so. The ability to choose one’s own caregivers was the most common response (see Table 19). Increased feelings of independence, dignity, and control over their care were the next most common responses, followed closely by the ability to purchase medicines.⁶ A number of clients said being able to obtain the right kind of PAS or buy personal items related to personal care and health had improved their lives the most.

I. CONCLUSION

On average, people in this early cohort of treatment group members were in relatively poor health and had high levels of functional impairment. Despite these limitations, one-third of the cohort members lived alone and one-quarter had no unpaid, informal caregivers. Most cash recipients (86 percent) used the monthly allowance to hire a caregiver on their own, and most did so within the first four months after enrolling. A majority of paid caregivers were relatives. Paid caregivers helped with all aspects of their clients’ care, although less often with routine medical care than with personal care or with household and community activities.

Clients were highly satisfied with the care they received. Still, substantial proportions said they needed more help than they were receiving. A substantial proportion disenrolled early (before the four-month interview). Nonetheless, 93 percent of clients, including disenrollees, would recommend IndependentChoices to others. Eighty-two percent said the monthly allowance had improved their lives, often by allowing them to choose their caregivers or buy medicine, or by enhancing their feelings of independence.

Our final analysis of data from the nine-month interviews will be based on a much larger sample. It will compare the experiences of randomly assigned treatment and control group members to see how beneficiaries receiving the monthly allowance and other IndependentChoices services fared relative to those receiving PAS through the traditional Medicaid program.

NOTES

1. Cash and Counseling is a national demonstration jointly funded by The Robert Wood Johnson Foundation and the U.S. Department of Health and Human Services, Office of the Assistant Secretary for Planning and Evaluation. The national program office for Cash and Counseling is the Center on Aging at the University of Maryland.

2. For the sake of brevity, this paper will refer to all survey respondents as clients of the Cash and Counseling program even though proxies sometimes responded on their behalf.

3. After data on the full sample are available for both groups, we will conduct an impact analysis comparing survey responses for all treatment and control group members. The impact analysis will also compare Medicaid and Medicare enrollment and claims data.

4. We calculated the response rate as of March 5, 2000. It excludes about 70 sample members for whom a completed survey was still being pursued. The response rate equals the number of respondents who completed interviews divided by the estimated number of eligible sample members. (The only ineligible sample members are those who die within one month of enrollment. We have no such cases thus far.)

5. Because we asked clients to identify all the hiring methods they used while trying to hire a caregiver, the percentages sum to more than 100 percent.

6. In Arkansas, Medicaid routinely covers three prescription drugs per month. At a physician’s request, the limit may be extended to as many six prescriptions. Under Arkansas’s demonstration rules, Cash and Counseling clients may use the monthly allowance for personal care or other medical- or personal assistance-related items or services not already covered by Medicaid.