Money Follows the Person: Change in Participant Experience

THE NATIONAL EVALUATION OF THE

MONEY FOLLOWS THE PERSON (MFP) DEMONSTRATION GRANT PROGRAM

REPORTS FROM THE FIELD

Number 6 • May 2011

Money Follows the Person: Change in Participant Experience During the

First Year of Community Living

By Samuel E. Simon and Matthew R. Hodges

Introduction

The Money Follows the Person (MFP) Demonstration, established by Congress through the

Deficit Reduction Act of 2005 (DRA), provides state Medicaid programs the opportunity to help

transition into the community Medicaid beneficiaries living in long-term care institutions. The

MFP program provides states with enhanced federal matching rates for spending on home- and

community-based services (HCBS) provided to program participants and provides funding for

associated administrative costs. MFP grantees, in turn, provide enhanced community services to

participants during their 365-day period of program participation. In 2007, the Centers for Medi-

care & Medicaid Services (CMS) awarded grants to, and has since overseen the implementation

of MFP programs in 29 states and the District of Columbia.

CMS awarded another 13 grants in

February 2011; these states are currently getting the details of their programs approved and start-

ing program operations.

The belief that many in institutional care would prefer to live in the community forms part of the

basis for the MFP demonstration program. This report presents the quality-of-life experiences of

803 MFP participants who transitioned to community living between January 2008 and December

2009 and responded to grantees’ administration of pre-transition and one-year post-transition

surveys. We specifically examine how reported quality of life changed after participants transi-

tioned to community living. We find that, after one year of community living, participants reported

significantly higher quality of life compared with life in institutional settings, as measured through

a variety of questions. These include questions designed to assess global satisfaction with life, sat-

isfaction with care received, and satisfaction with where participants lived. Participants reported

the largest improvement in satisfaction with their living arrangements. These findings are consis-

tent across each target population. The report concludes with a discussion of the implications of

these findings for the demonstration and for other transition programs.

One additional state received a grant but has not implemented its MFP program. Hereafter, we refer to the original 30 grantees, includ-

ing the District of Columbia, as grantee states.

ABOUT THE MONEY FOLLOWS THE PERSON DEMONSTRATION

The MFP demonstration, first authorized by Congress as part of the 2005 DRA and then extended by the 2010

Patient Protection and Affordable Care Act, is designed to shift Medicaid’s long-term care spending from

institutional care to HCBS. Congress has now authorized up to $4 billion in federal funds to support a twofold

effort by state Medicaid programs to (1) transition people living in nursing homes and other long-term care

institutions to homes, apartments, or group homes of four or fewer residents; and (2) change state policies so

that Medicaid funds for long-term care services and supports can “follow the person” to the setting of his or

her choice. MFP is administered by CMS, which initially awarded MFP grants to 30 states and the District of

Columbia (although one state has yet to implement its program). The first states launched their MFP transi-

tion programs in late 2007, and Congress has authorized the demonstration through 2016. CMS contracted

with Mathematica Policy Research to conduct a comprehensive evaluation of the MFP demonstration and

report the outcomes to Congress.

An operating premise of the MFP program is that

many Medicaid beneficiaries who live in institutions

would rather live in the community and that success-

ful transition to the community will improve the qual-

ity of their lives. Although the welfare of people who

transition from institutional settings into less restric-

tive community-based care is a primary concern of

long-term care providers, beneficiaries, and their

advocates, little is known about how this transition

affects the quality of their lives. This report begins

to shed light on this issue and test the premise that

community living enhances the quality of life of those

who need long-term services and supports by examin-

ing how participants’ reported quality of life changed

after a year of community living.

Since 2007, 30 grantees have established MFP pro-

grams. Within specified parameters, each grantee has

established a unique set of goals for transitioning target

populations—such as which beneficiaries will be the

focus of their program and how many of each category

of beneficiaries will be transitioned—and other related

objectives to measure success in program implementa-

tion. As one component of their programs, grantees

survey potential transition candidates using a standard-

ized instrument to assess participant quality of life in

terms of satisfaction, access to personal care, com-

munity integration, respect and dignity, and degree of

choice and control. Grantees use the MFP-Quality of

Life survey instrument (MFP-QoL) to collect informa-

tion directly from participants. The Data and Methods

box at the end of this report describes this instrument

and its administration (Sloan and Irvin 2007). Grantees

survey MFP participants annually for two years follow-

ing the initial transition to the community to obtain the

data needed to assess the change in their quality

of life.

This report provides preliminary results that show

how participant quality of life changed over time and

explores policy implications for the ongoing MFP pro-

gram, its participants who have transitioned and those

yet to do so, and other transition programs designed with

comparable goals and objectives. This report describes

changes for specific subpopulations, including those

who transitioned from an intermediate care facility for

the mentally retarded (ICF-MR), those who are aged

(defined as age 65 or older) and transitioned from a

nursing facility, and those who are younger than 65 and

also transitioned from a nursing facility.

Subsequently,

we refer to these target populations in the report as indi-

viduals with intellectual disabilities (ID), the aged, and

individuals with physical disabilities (PD), respectively.

Focusing on target populations reflects how many MFP

programs operate in practice. Grantees typically tailor

their transition programs to the target population, part-

nering with different organizations (such as Area Agen-

cies on Aging, Centers for Independent Living, and the

state agency for developmental disabilities for individu-

The first follow-up survey is to be conducted approxi-

mately 11 months after transition; the second follow-up

survey is to be conducted 24 months after transition. When

necessary, survey respondents may use either the interpre-

tive assistance of another or a proxy respondent to provide

answers. See the Data and Methods box for more details re-

garding the survey instrument.

A fourth category—those who transitioned from institu-

tions for mental diseases (IMDs)—represents a much smaller

and more specialized segment of the MFP program’s partici-

pants, and not enough observations for this target population

were available for this report.

als with intellectual disabilities), using different target-

ing strategies, and placing different target populations

in different HCBS waivers (Lipson and Williams 2010).

We examined how the following broad dimensions of

participants’ experiences changed between the time they

were in institutional care and one-year post-transition:

(1) satisfaction with life (or life satisfaction), (2) quality

of care, and (3) community life.

Data for this report include survey and administrative

data submitted to CMS through November 2010, repre-

senting surveys conducted through September 2010. We

included only participants with both pre-transition (or

baseline) and follow-up survey data.

All pre-post transi-

tion changes reported as statistically significant were

tested using paired t-tests and are significantly different

at p < .01. We examined change in participant quality of

life regardless of enrollment or institutionalization status

at the time of the one-year follow-up survey. Subse-

quent reports on participant experience will examine

how changes in quality of life vary with MFP program

components and individual characteristics (such as age,

physical functioning, and health care utilization).

An earlier examination of MFP participant experi-

ence before transition identified several key findings

that affect the analysis of how quality of life changes

after the initial transition to the community (Irvin

et al. 2010).

First, most MFP participants reported

Baseline surveys were conducted between January 2008

and December 2009. To ensure that follow-up surveys cap-

tured experience at one year, follow-up surveys conducted

between 8 and 16 months after the baseline survey were in-

cluded in the study sample.

The earlier findings were based on analyses of responses

to the initial or baseline survey, conducted through December

2009, and included 1,890 MFP participants from 25 of 30

grantee states.

high levels of happiness with the way they lived their

lives and the care they received in the weeks and days

before transitioning. In addition, before transition,

participants with intellectual disabilities reported

relatively high levels of life satisfaction (74 percent)

compared with aged participants and participants

with physical disabilities (57 and 56 percent, respec-

tively). These findings suggest that the transition to

community living will have differential effects on

the quality of life of different target populations and

that some participants entered community living with

relatively high levels of satisfaction with their lives.

For example, because of their relatively high level

of baseline satisfaction with life, participants with

intellectual disabilities may be less likely than other

participants to show improvement.

Characteristics of the Study Sample

In the sample of MFP participants represented in this

report, aged participants (17 percent), participants with

physical disabilities (27 percent), and participants with

intellectual disabilities (28 percent) comprised nearly

three-quarters of the study sample (Table 1). The rest

represented those transitioning from another site of

institutionalization (2 percent) or those who could not

be classified due to missing data (26 percent). Data

from 22 of the 30 grantees are included, with data from

Connecticut, Missouri, Ohio, and Oregon representing

54 percent of the analytic sample.

Most of the sample (60 percent) was younger than

age 65, which is consistent with the overall profile

of MFP participants (Lipson and Williams 2010).

Approximately 21 percent of the study sample are

those age 65 and older (Table 2). Children and young

adults (those younger than 21) represented a small

percentage (2 percent), but for the balance of the

TABLE 1. STUDY SAMPLE, BY TARGET POPULATION

Target Population Number Percentage

Total

803

100.0

Aged

138 17.2

Participants with physical disabilities

217 27.0

Participants with intellectual disabilities

228 28.4

Other

14 1.7

Unknown

206 25.7

Source: Mathematica analysis of MFP-QoL survey and Program Participation data files submitted through November 2010,

representing transitions through December 2009 and 1-year follow-up surveys through September 2010.

TABLE 2. CHARACTERISTICS OF MFP PARTICIPANTS IN STUDY SAMPLE

Characteristics Number Percentage

Total

803 100.0

Age Distribution at Time of Transition

<21

18 2.2

21-44

152 18.9

45-64

313 39.0

65-74

88 11.0

75-84

49 6.1

>=85

33 4.1

Unknown

150 18.7

Sex

Female

399 49.7

Male

395 49.2

Unknown

9 1.1

Reinstitutionalized

Yes

54 6.7

749 93.3

Source: Mathematica analysis of MFP-QoL survey and Program Participation data files submitted through November 2010,

representing transitions through December 2009 and 1-year follow-up surveys through September 2010.

Note: Reinstitutionalizations of 30 days or longer are reported in the MFP Program Participation data.

sample, age was not reported by the grantee. Table 2

also shows the percentage of the population who were

ever reinstitutionalized.

Does Quality of Life Improve After the

Transition to Community Living?

One year after transition to the community, MFP par-

ticipants reported improvement in the quality of their

lives across all domains considered. Table 3 displays

the magnitude of improvement overall and by target

population for key indicators across six domains of

participant quality of life.

Participants reported the

largest improvement in satisfaction with their living

arrangements, with satisfaction among participants

with physical disabilities increasing by 50 percentage

Approximately 5 percent of participants in the sample

were surveyed while institutionalized. Grantees are to con-

duct both follow-up surveys, regardless of the participant’s

institutional status.

Areas of choice and control also demonstrated improve-

ments of similar relative magnitude. However, a different

unit of measurement was used: average number of areas of

choice and control reported. Figure 8 shows information for

choice and control.

points between pre-transition and one-year post-tran-

sition to community living.

How Does Transition to the Community Affect

MFP Participants’ Quality of Life?

The MFP-QoL survey reflects the concept that an

individual’s quality of life is multidimensional and is

a function of overall life satisfaction, quality of care

received, and community life.

Overall Life Satisfaction. For the MFP program

to be successful, satisfaction with life must be

maintained or improved in community-based

settings. As Figure 1 shows, MFP participants

across all target populations reported a significant

increase in their satisfaction with the way they lived

their lives.

Among all target populations, nearly

60 percent of participants reported being satisfied

with the way they lived their life while still in

institutional care. This percentage increased to 81

The survey question was: “Taking everything into con-

sideration, during the past week, have you been happy or

unhappy with the way you live your life?”

All Participants Aged PD ID

(N = 803) (N = 138) (N = 217) (N = 228)

Target Population

 Pre-Transition  Year 1

100

Percentage of Participants Reporting

Satisfaction with Life

TABLE 3. SUMMARY OF FINDINGS, BY TARGET POPULATION

Domain Indicator

Total Aged PD ID Other/ Unknown

Satisfaction with life

++ ++ ++ + ++

Unmet personal care needs

+ + + +

Respect and dignity

++ ++ ++ + ++

Satisfaction with living arrangements

++++ ++++ +++++ ++ +++++

Community integration

+ ++ + + +

Mood status

+ +

803 138 217 228 220

Source: Mathematica analysis of MFP-QoL survey and Program Participation data files submitted through November 2010,

representing transitions through December 2009 and 1-year follow-up surveys through September 2010.

For aspects of participant experience where a single, key question can be identified, that question is used to represent the

domain. For example, we use the question assessing whether participants felt “sad or blue” to represent the domain of mood

status.

+ Indicates improvement of 10 to 19 percentage points.

++ Indicates improvement of 20 to 29 percentage points.

+++ Indicates improvement of 30 to 39 percentage points.

++++ Indicates improvement of 40 to 49 percentage points.

+++++ Indicates improvement of 50 or more percentage points.

Figure 1. Percentage of MFP Participants

Reporting Satisfaction with Life,

Pre-Transition and Post-Transition

Source: Mathematica analysis of MFP-QoL survey and

Program Participation data files submitted through

November 2010, representing transitions through

December 2009 and 1-year follow-up surveys

through September 2010.

percent one year after the transition to community-

based care.

The high rate of satisfaction with life after one year in

the community is attributable to broad improvement in

satisfaction. That is, when the analysis was restricted

to those who were not satisfied with their lives while

in institutional care, 72 percent reported being satis-

fied with life at the one-year follow-up. Despite these

favorable outcomes, some MFP participants became

less satisfied with their lives. Among those who were

satisfied with their lives while in institutional care, 13

percent reported they were no longer satisfied a year

Analyses of proxy respondents were also conducted

to determine the influence proxies might have had on the

results. Unlike earlier analyses of the quality of life of MFP

participants before they transitioned to community living,

the use of a proxy respondent was not associated with sat-

isfaction with life (Irvin et al. 2010). For the study sample

presented in this report, proxy respondents provided quality

of life information for 23 percent of pre-transition interviews

and 16 percent of post-transition interviews. At both time

points, ratings of satisfaction with life did not vary by the

type of respondent (the participant or a proxy). This find-

ing contrasts with our earlier examination, which found that

proxies reported higher rates of satisfaction with life than

what MFP participants themselves reported. This contrast

between the different analyses is most likely related to how

the study samples were defined.

later. Among those experiencing a decline in their

satisfaction, 13 percent were in institutional care at the

time of the one-year follow-up survey.

Satisfaction with life increased across all target

populations. The percentage of aged and those with

physical disabilities reporting overall satisfaction

with life increased by 26 and 23 percentage points,

respectively, at the one-year assessment. Although

participants with intellectual disabilities had the

smallest increase over time, a greater percentage were

satisfied with their lives at baseline when they were

still in institutional care, compared with nursing

home residents.

The finding that about 8 of 10 participants reported

satisfaction with the way they live their life one year

after transition compares favorably with the rate

reported by Beauchamp et al. (2006) in their report on

overall life satisfaction for participants in the Program

of All-Inclusive Care for the Elderly (PACE) and

HCBS users (2008). Using an item identical to the one

in the MFP-QoL survey, Beauchamp and colleagues

surveyed PACE participants and HCBS users between

18 months and five years post-enrollment and reported

overall satisfaction rates of 71 percent for HCBS

recipients and 74 percent for PACE participants.

We conducted additional analyses of specific sub-

groups to determine the relationship between satis-

faction with life and reinstitutionalization. Overall, 5

percent of the study sample had returned to institu-

tional care at the time of their one-year follow-up sur-

vey.

Among those reinstitutionalized, only 49 percent

indicated they were satisfied with the way they lived

their life a year after their initial transition to commu-

nity living, approximately 30 percentage points lower

than what is observed in the overall study sample.

Satisfaction with life among reinstitutionalized popula-

tions ranged from 43 percent for aged participants to

67 percent for participants with intellectual disabilities.

These rates of satisfaction among reinstitutionalized

participants were lower than the pre-transition rates for

each target population.

Quality of Care. Some stakeholders express

concern that a potential consequence of community

living is poorer-quality care due to diffusion of

accountability among a wider range of community-

Nearly seven percent of the analytic sample were rein-

stitutionalized during their first year of MFP participation.

based providers. To assess care quality, the quality-

of-life survey uses reported satisfaction with

care, unmet need for personal care assistance, and

treatment by providers.

For all groups of participants, after the first year

of community living, all three areas of care quality

improved from pre-transition levels. MFP participants

were more satisfied with the care they received, had

fewer reports of unmet personal care needs, and more

reported their caregivers treated them with respect

and dignity.

Satisfaction with Care. Although nearly three-quarters

of participants (71 percent) were satisfied with the help

they received in an institutional setting, an even larger

proportion (90 percent) were satisfied with the assis-

tance they received in the community. Satisfaction with

care increased across all target populations.

Unmet Need for Personal Care Assistance. A promi-

nent feature of institutional care is assistance with

activities of daily living (ADLs) and instrumental

activities of daily living (IADLs). Although a com-

munity residence offers greater freedom, obtaining

personal care assistance in the community may be

difficult if paid and unpaid caregivers who provide

essential assistance with ADLs and IADLs are not

always available when needed. Therefore, respondents

to the MFP-QoL survey were asked to report unmet

needs in the areas of (1) bathing, (2) meal preparation,

(3) medication management, and (4) toileting.

Despite concerns that the transition to a community

setting could lead to an unintended decrease in meeting

personal care needs, we found evidence to the contrary.

Figure 2 shows that the percentage of participants

reporting one or more unmet care needs significantly

decreased between the pre- and post-transition periods

(15 to 4 percent), and this finding was consistent across

target populations.

To further assess the post-transition unmet care needs

of MFP participants more generally, the survey asked

respondents about post-transition support. Nearly

half of all participants (47 percent) indicated they had

To assess satisfaction with care, the survey asks: “Tak-

ing everything into consideration, during the past week, have

you been happy or unhappy with the help you get with things

around the house or getting around your community?”

Unmet needs are defined as ever going without a par-

ticular activity because of a lack of assistance.

Percentage of Participants Reporting

One or More Unmet Care Needs

All Participants Aged PD ID

(N = 803) (N = 138) (N = 217) (N = 228)

Target Population

 Pre-Transition  Year 1

100

Percentage of Participants Reporting Being

Treated with Respect and Dignity

All Participants Aged PD ID

(N = 803) (N = 138) (N = 217) (N = 228)

Target Population

 Pre-Transition  Year 1

Figure 2. Percentage of MFP Participants

Reporting One or More Unmet

Care Needs, Pre-Transition and

Post-Transition

Source: Mathematica analysis of MFP-QoL survey and

Program Participation data files submitted through

November 2010, representing transitions through

December 2009 and 1-year follow-up surveys

through September 2010.

spoken with a case manager or support coordinator

about the need for special equipment or changes to

the home that would make life easier.

Of these, most

(85 percent) said they received the desired equipment

or changes. When asked whether they needed more

help with tasks around the house, such as cooking and

cleaning, fewer than one in six (15 percent) acknowl-

edged such a need. In addition, more than one-third of

respondents (37 percent) indicated they received help

from family and friends with such tasks, suggesting

that informal support outside the MFP program exists

and provides supplemental benefits.

Treatment by Providers. Respectful treatment by paid

caregivers is a key component of quality of life for any

recipient of institutional or home- and community-based

long-term care. In institutional settings such as nursing

homes, CMS maintains standards of care that focus on

treating residents respectfully and with dignity. These

safeguards, however, are less systematic and enforceable

for community-based care. Therefore, MFP participants’

Special equipment was defined as items such as wheel-

chairs, canes, vans with lifts, and automatic door

openers.

assessments of how they are treated once they live in the

community and receive services there are a vital compo-

nent in measuring quality of life.

Despite the concern about reduced safeguards, Figure 3

shows that MFP participants experienced a significant

improvement in their treatment, increasing from 67 per-

cent reporting treatment with respect and dignity during

the pre-transition period to 92 percent a year after they

transitioned to the community.

Findings also demon-

strated significant increases for each component of the

measure independently, though neither component alone

(82 to 96 percent for being treated well and 79 to 96 per-

cent for being listened to) demonstrated the magnitude

of improvement that the combined measure did.

Findings also revealed that reported physical abuse (an

optional question designed to detect mistreatment by

staff) decreased significantly post-transition (2 percent

The results combine responses to two questions asked

of MFP participants who have caregivers who help them with

everyday activities: “Do the people who help you treat you

the way you want them to?” and “Do the people who help

you listen carefully to what you ask them to do?”

Figure 3. Percentage of MFP Participants

Reporting Treatment with Respect

and Dignity, Pre-Transition and

Post-Transition

Source: Mathematica analysis of MFP-QoL survey and

Program Participation data files submitted through

November 2010, representing transitions through

December 2009 and 1-year follow-up surveys

through September 2010.

All Participants Aged PD ID

(N = 803) (N = 138) (N = 217) (N = 228)

Target Population

 Pre-Transition 

100

Percentage of Participants Reporting

Satisfaction with Living Situation

of participants reported physical abuse, compared with 6

percent pre-transition).

Among target populations, par-

ticipants with intellectual disabilities and aged participants

reported significantly less abuse after transitioning (1 and

2 percent, respectively) compared with the rates reported

while institutionalized (5 and 8 percent, respectively).

Community Life. A fundamental expectation of

transitioning from institutional to community living

is that participants will have enhanced community

experiences. The MFP-QoL survey measures

satisfaction with living arrangements, community

engagement, work status, and degree of choice

and control as proxies for participants’ quality of

community life.

Satisfaction with Living Arrangements. Most MFP

grantees report difficulty finding appropriate hous-

ing for participants (Denny-Brown et al. 2011). The

environments into which MFP participants transition

vary, from homes and apartments (sometimes with

family) to group homes of four or fewer residents. As

Figure 4 shows, nearly all MFP participants (94 per-

cent) reported liking where they lived nearly one year

after transitioning, an increase large both in magnitude

(compared with the 54 percent who reported liking

their living arrangement while institutionalized) and

significance. Improvement was significant for each

target population. Likewise, improvement was wide-

spread; among all participants not satisfied with their

living situation in an institutional setting, 92 percent

reported being happy at the one year follow-up (infor-

mation not shown). Just 5 percent reported a decline

in their satisfaction with living arrangements, and 18

percent of these participants were in institutional care

at the time of the one-year follow-up survey.

Community Integration and Inclusion. Integration

and inclusion in one’s community can help minimize

the potential for depression and loneliness that some

people may experience transitioning from group set-

tings to independent living. MFP participants report

a high level of community integration and inclusion

after transitioning to community living. For example,

interactions with family and friends remained high

after participants left institutional settings. Further-

more, most MFP participants reported an ability to see

friends and family when they chose to, with the pro-

portion doing so post-transition (90 percent) slightly

This question was answered by 481 respondents pre-

transition and 470 respondents post-transition.

Figure 4. Percentage of MFP Participants

Reporting Satisfaction with Living

Situation, Pre-Transition and

Post-Transition

Source: Mathematica analysis of MFP-QoL survey and

Program Participation data files submitted through

November 2010, representing transitions through

December 2009 and 1-year follow-up surveys

through September 2010.

greater compared with the period before transition (85

percent). Nearly all participants reported an ability to

get to needed places such as work, shopping, or the

doctor’s office pre- and post-transition (89 and 95

percent, respectively). Only participants with physical

disabilities experienced a significant increase in their

ability to get around the community (93 percent post-

transition compared with 81 percent pre-transition).

A substantial change in community integration associ-

ated with transition was participants’ ability to partici-

pate in activities outside their homes. Figure 5 shows

that half of all MFP participants (50 percent) were

unable to participate in community activities while in

institutional care, but this proportion declined to 37

percent one year after the initial transition to commu-

nity living. This limitation generally varied among the

target populations, with the proportion of participants

with physical disabilities who were unable to partici-

pate in community activities after transition to com-

munity living (44 percent) exceeding the rate for aged

participants (37 percent) and participants with intellec-

tual disabilities (25 percent).

Year 1

All Participants Aged PD ID

(N = 803) (N = 138) (N = 217) (N = 228)

Target Population

 Pre-Transition  Year 1

Percentage of Participants Unable to Do Things

They Want to in the Community

Working for pay

Not working for pay,

with desire to work for

pay

Neither working for pay

nor willing to work for

Other/Unknown

pay

Figure 5. Percentage of MFP Participants

Reporting Barriers to Community

Integration Pre-Transition and

Post-Transition

Source: Mathematica analysis of MFP-QoL survey and

Program Participation data files submitted through

November 2010, representing transitions through

December 2009 and 1-year follow-up surveys

through September 2010.

Assessed through the question: “Is there anything you want

to do outside of your home that you cannot do now?”

We assessed post-transition outcomes related to paid

employment and work on a volunteer basis.

Nearly

one-fifth of all MFP participants (19 percent) reported

working for pay (Figure 6). As Figure 7 shows, par-

ticipants with intellectual disabilities represented the

greatest proportion of paid workers (72 percent). Fewer

than one in ten (9 percent) participants reported doing

volunteer work after transition (data not presented).

One-fifth of participants (20 percent) did not report

working for pay but indicated a desire to do so. Partici-

pants with physical disabilities represented the greatest

proportion not working but willing to work for pay or on

a volunteer basis (44 and 35 percent, respectively) (data

not presented).

Choice and Control. Program participants should

expect increased autonomy after they transition to

community living. The MFP-QoL survey assesses six

Questions related to employment and volunteer status

were administered at the one-year follow-up survey only.

Figure 6. MFP Participants’ Paid Work Status

After One Year of Community

Living (N = 787)

61%

19%

20%

Source: Mathematica analysis of MFP-QoL survey and

Program Participation data files submitted through

November 2010, representing transitions through

December 2009 and 1-year follow-up surveys

through September 2010.

Figure 7. MFP Participants Who Worked for Pay

After One Year of Community Living

(N = 148)

23%

72%

Source: Mathematica analysis of MFP-QoL survey and

Program Participation data files submitted through

November 2010, representing transitions through

December 2009 and 1-year follow-up surveys

through September 2010.

Aged participants did not report working for pay.

areas of choice and control: being able to go to bed

when one desires, the ability to be alone when one

chooses, the ability to eat food of one’s choice and

when one chooses, and the ability to use the tele-

phone or watch television when one chooses.

As Figure 8 shows, MFP participants’ choice and con-

trol increased significantly between the pre-transition

All Participants Aged PD ID

(N = 803) (N = 138) (N = 217) (N = 228)

Target Population

 Pre-Transition  Year 1

6.0

5.0

4.0

3.0

2.0

1.0

0.0

Reported Areas of Choice and Control

3.5

4.9

3.7

4.7

3.7

5.0

3.2

4.4

Figure 8. Average Number of Areas of Choice

and Control Reported by MFP Participants,

Pre-Transition and Post-Transition

Source: Mathematica analysis of MFP-QoL survey and

Program Participation data files submitted through

November 2010, representing transitions through

December 2009 and 1-year follow-up surveys

through September 2010.

Note: The MFP-QoL survey assesses six areas of choice

and control: being able to go to bed when one

desires, the ability to be alone when one chooses,

the ability to eat food of one’s choice and when one

chooses, and the ability to use the telephone or watch

television when one chooses.

period and one year post-transition (from an average of

3.5 areas of choice and control to 4.9 areas). Moreover,

results revealed that autonomy among each of the six

individual measures of choice and control increased

significantly, although most gains were modest. Trends

in increased choice and control were similar across the

different target populations.

Implications of Participant Quality of

Life One Year After Transition to

Community Living

This early assessment of MFP participants’ experi-

ence a year after their initial transition to community

living suggests that participants experienced significant

improvement in their quality of life. Should the trends

reported here continue throughout the demonstra-

tion, then most MFP participants can be expected to

experience an enhanced quality of life once outside

the institutional setting. Other preliminary findings

show that MFP had a broad effect across all target

populations, with each quality-of-life domain show-

ing statistically meaningful change from pre-transition

levels. These early findings also suggest that MFP is

doing more than simply facilitating transitions to the

community, as participants reported increased levels

of community integration, as well as reduced levels of

unmet care needs and expanded choice and control in

community settings. Global ratings of quality of life

increased significantly and, given our inclusion of the

small but significantly less satisfied institutionalized

population, may be a slight underestimate of satisfac-

tion among participants who successfully transition to

the community.

We observed the largest relative improvements for

satisfaction with living arrangement and the percentage

of participants reporting any unmet care needs; both

measures improved by 74 and 73 percent, respec-

tively. Although mood status improved significantly

at one-year post-transition, this domain showed the

least improvement; still, the percentage of participants

reporting they felt sad or blue fell 19 percent between

pre-transition and one-year follow-up. However, in

spite of this improvement, more than one-third of MFP

participants (38 percent) reported feeling sad or blue

over the past seven days following transition to the

community.

The paradoxical finding that 8 of 10 participants indi-

cated they were satisfied with the way they lived their

life post-transition, yet a substantial minority reported

depressed mood, warrants further attention. It is possi-

ble that respondents overstate their satisfaction ratings

or that their reported mood levels are biased in some

way. Regardless, the overall magnitude of reported

depressed mood suggests that providers and HCBS

programs should carefully monitor mood for partici-

pants who make the transition from institutional care

to community. Although a participant may be happy to

move to the community, the transition to community

living has the potential for isolation or for community

living not matching a participant’s expectations. Future

analyses of the quality-of-life data collected by MFP

grantees will include additional assessments of mood

and its relationship to critical outcomes such as satis-

faction and reinstitutionalization.

Our findings may have been influenced by the avail-

ability of informal support networks for MFP partici-

pants. Beyond the linkages to the formal services MFP

provides, participants transitioning to a community set-

ting rely upon informal social and care supports. The

contribution of such informal networks to enhanced

quality of life is unknown, but their economic value

within the U.S. health care system is estimated to

exceed those of the formal health care and nursing

home care sectors combined (Arno et al. 1999).We

observed that about one-third of sample participants

received informal support from family and friends. The

extent of these informal supports, exogenous to the

MFP program, may play an important role in maintain-

ing participant quality of life. Monitoring the provision

of such services will increase in importance following

the 365-day period of MFP participant service provi-

sion (that is, when formal MFP supports are no longer

available). Family and friends play a vital role in this

regard, but so do other forms of social networks, such

as faith-based institutions and congregations, commu-

nity centers, and volunteer organizations.

Identification of, and access to, appropriate hous-

ing is a frequently reported barrier for many MFP

grantees (Denny-Brown et al. 2011). However, our

findings indicate that, after housing is secured,

respondents are overwhelmingly satisfied with their

living arrangements (94 percent). This finding also

implies that grantees have largely been successful in

meeting the housing needs of the participants repre-

sented in this analysis.

About one in five MFP participants reported working

for pay during the first year living in the community,

although employment varied widely by target popula-

tion. Overall, 48 percent of all participants who wanted

to work for pay reported doing so.

Nearly half of all

participants with an intellectual disability reported paid

work experience, while no aged participants and very

few participants with physical disabilities (4 percent)

worked for pay. Among those not working, the desire

to work was most common among participants with

physical disabilities.

Although the findings of this report are largely positive,

we note several areas where participant experience can

be improved. First, many participants who are not cur-

rently working, particularly participants with physical

disabilities, would like to be. The high level of interest

Calculated as the number working divided by the num-

ber working plus the number not working but reporting that

they wanted to work.

in, and low reported rates for, paid work, represents an

important opportunity for grantees to further enhance

the integration of MFP participants into their commu-

nities. Second, about a third of all participants report

barriers to community integration, with more than 4

in 10 participants with physical disabilities reporting

they were unable to participate in community activities.

Third, 6 percent of aged and participants with physi-

cal disabilities reported unmet care needs. Because the

survey focuses on assistance with such tasks as bathing,

meal preparation, medication management, and toilet-

ing, not receiving assistance with these activities greatly

diminishes quality of life and may introduce unneces-

sary risk for the participant. Finally, depressed mood

continues to be reported for a noteworthy portion of

participants, and more attention to mood status may be

necessary to ensure that the participant has a successful

transition to community living.

Study Limitations

Several important limitations of our analysis warrant

consideration and caution concerning the overall pos-

itive findings presented in this report. The informa-

tion in this report should be considered preliminary

before our findings can be taken as representative of

the program, as our findings need to be replicated

for a larger proportion of participants, and baseline

characteristics of participants need to be controlled

for in the analyses.

Compared with recent reports on the number and types

of MFP participants who transitioned through MFP, our

analytic sample under-represents aged and physically

disabled participants (Lipson and Williams 2010). The

sample reflects participants who had data submitted on

their behalf and where a linkage between the pre- and

post-transition surveys could be established and both

could be linked with administrative data records. In

addition, the data set includes participants from 22 of

the 30 grantees. Given that MFP transitioned approxi-

mately 5,600 participants through December 2009,

our study sample represents fewer than one in six of

all transitions that could have an associated follow-

up one-year assessment by September 2010, the last

month of data included in this report. Not having data

on these participants may bias the results if excluded

participants were systematically different from those

in our sample. To assess differences between this

Future reports will address these issues.

group of participants and participants not represented

in this analysis, we compared pre-transition ratings of

global satisfaction for the participants in this analy-

sis with those of a larger group of MFP participants

not included in this study (N = 1,779) and found that

baseline global satisfaction scores were not signifi-

cantly different. However, we caution that our findings

will require replication in a larger group of participants

before we consider them representative.

Another important data limitation was lack of institu-

tionalization site and age data for approximately one-

quarter of the respondents. These variables are missing

on grantee-submitted MFP Program Participation Data

files; however, they will be available through follow-

up with the grantees and linking these data to Medicaid

eligibility records, which will be conducted at later

stages in our analysis.

The method of survey administration is also a poten-

tial source of bias for these data. Because grantees

administer the MFP-QoL survey, and surveys are often

conducted by MFP transition coordinators, consumers

may feel compelled to overstate satisfaction if they felt

a need to provide positive information or believed that

reporting problems could result in negative outcomes

for themselves. Although there is no evidence that this

occurred, this phenomenon cannot be ruled out as a

bias in the data in terms of absolute values. However,

if the bias is similar between the pre- and post-transi-

tion surveys, the change in improved outcomes should

not be affected.

Another possible confound to our findings is that quality

of life and ratings of participant experience are, by their

very nature, subjective entities. Participant expectation

of transition at the time of the pre-transition survey may

color ratings of participant experience. However, given

the significant increases in participant-rated quality

of life at follow-up, this confound does not appear to

have a strong effect on our findings. Should it exist, the

changes we can document with these data may under-

state the true change in quality of life.

We acknowledge this analysis excludes a range of

unmeasured program and individual factors that are

likely to affect participants’ first-year experience. For

example, we plan to explore program characteris-

tics such as model of caregiver employment (agency

versus self-direction). Similarly, we are interested in

understanding how the type and volume of HCBS

received affects participants’ experience. Furthermore,

grantees have identified specific activities to rebal-

ance long-term care by expanding the availability of

HCBS; providing transition services (for example, case

management, housing assistance, or one-time transition

expenditures); or providing other innovations, such

as investing in assistive technology (Denny-Brown et

al. 2011). Finally, we plan to control for differences in

participant characteristics at discharge using Minimum

Data Set data for participants who transition from nurs-

ing homes. Many of these activities and characteristics

are likely to affect participant quality of life and will be

explored in future reports.

Acknowledgments

This research was conducted by Mathematica under

contract with CMS (HHSM-500-2005-00025I). The

authors wish to thank Carol Irvin, Randall Brown, and

Robert Schmitz for their insightful comments on earlier

drafts. Thanks also go to Chris Rodger, Dean Miller,

Greg Bee, and Amanda Hakanson for programming

assistance. The authors also gratefully acknowledge

CMS, the MFP grantee states for providing the data

used in this report, and most of all, the MFP-QoL

survey respondents.

References

Arno, Peter S., Carol Levine, and Margaret M. Mem-

mott. “The Economic Value of Informal Caregiving.”

Health Affairs, vol. 18, no. 2, March/April 1999, pp.

182–188.

Beauchamp, Jody, Valerie Cheh, Robert Schmitz, Peter

Kemper and John Hall. “The Effect of the Program of

All-Inclusive Care for the Elderly (PACE) on Quality.”

Final report submitted to the Centers for Medicare &

Medicaid Services. Princeton, NJ: Mathematica Policy

Research, February 12, 2006.

Denny-Brown, Noelle, Debra Lipson, Christal Stone

and Jessica Ross. “Money Follows the Person Demon-

stration: Overview of State Grantee Progress, Janu-

ary – June 2010.” Report submitted to the Centers

for Medicare & Medicaid Services. Cambridge, MA:

Mathematica Policy Research, January 2011.

Irvin, Carol, Debra Lipson, Sam Simon, Audra Wen-

zlow, and Jeffrey Ballou. “Money Follows the Person

2009 Annual Evaluation Report.” Final report submit-

ted to the Centers for Medicare & Medicaid Services.

Cambridge, MA: Mathematica Policy Research,

September 10, 2010.

Lipson, Debra J., and Susan R. Williams. “The Money

Follows the Person Demonstration Program: A Profile

of Participants.” National Evaluation of the Money

Follows the Person (MFP) Demonstration Grant Pro-

gram, Reports from the Field No. 5. Washington, DC:

Mathematica Policy Research, December 2010.

Sloan, Matt, and Carol Irvin. “Money Follows the Per-

son Quality of Life Survey.” Prepared for the Centers

for Medicare & Medicaid Services. Washington, DC:

Mathematica Policy Research, 2007.

DATA AND METHODS

Data Sources

The statistics presented in this report are derived from two data sources: (1) MFP-QoL surveys, and (2) Program

Participation Data files. All MFP participant experience data are obtained through administration of the MFP-

QoL survey by grantees. The MFP-QoL was developed by Mathematica staff and is based largely on the Partici-

pant Experience Survey, with several items drawn from other instruments.

The MFP-QoL instrument captures

seven aspects of participants’ quality of life: (1) overall satisfaction, (2) satisfaction with living arrangements,

(3) unmet need for personal care needs, (4) respect and dignity, (5) choice and control, (6) community integration

and inclusion, and (7) mood status.

Data from the MFP-QoL were matched with MFP Program Participation Data files; both file types are submit-

ted by grantees quarterly. The participation files confirm participant eligibility status and provide information

on qualified institutional stays, qualified residence, and reinstitutionalization. This analysis used MFP-QoL and

participation data submitted to CMS through November 2010, representing transitions through December 2009

and 1-year follow-up surveys through September 2010. We included individuals’ baseline and one-year follow-up

survey records that could be matched with a record in the MFP Program Participation Data files.

To classify the MFP target populations, we used information on the qualified institution and age as reported at

the time of an individual’s initial transition to the community to construct MFP target group assignments for

enrollees (both are reported in the MFP Program Participation Data files). The aged were individuals 65 years

or older who transitioned from a nursing home; individuals with physical disabilities were younger than 65 who

transitioned from nursing homes; and individuals with intellectual disabilities were participants of any age who

transitioned from an ICF-MR. Individuals transitioning from any other type of institutional care were assigned to

“other,” and individuals without complete age or qualified institution data were categorized as “unknown.”

The analytic sample includes 803 people with institutional (baseline) and one-year follow-up MFP-QoL surveys

that could be matched with a participation record. To ensure that follow-up survey data reflected participant expe-

rience at one year, we included data for participants with a follow-up survey conducted between 8 and 16 months

after their institutional assessment. The sample includes participants from 22 of the 30 grantees, although data

from four states (Connecticut, Missouri, Ohio, and Oregon) make up 54 percent of the sample.

(continued)

Those instruments include ASK ME! Cash and Counseling, National Core Indicator Survey, Quality of Life Enjoyment and

Satisfaction Questionnaire—Short Form, and the Nursing Home Consumer Assessment of Health Plans Survey.

Grantees with no data in the sample are Delaware, Illinois, Indiana, Louisiana, Michigan, North Carolina, North Dakota, and

Virginia.

DATA AND METHODS (continued)

To assess participant experience, we selected key indicators for each domain on the MFP-QoL instrument. For

aspects of participant experience where a single, key indicator question can be identified, we selected that indica-

tor to represent the quality-of-life domain. For example, we used the item assessing whether participants felt sad

or blue to represent the domain of mood. In domains where we selected a single item to represent participant

experience, we also examined other related items to ensure consistency. For example, for the domain of choice

and control, information from all six questions assessing participant choice was summarized to yield a count of

areas of reported choice and control.

QoL Survey Administration

Grantees are responsible for survey administration, data entry, tracking, quality assurance, and transmission

of the data to CMS. The survey, which takes about 20 minutes to administer, consists of 41 questions and is

designed to be conducted in person and in a private setting. Grantees are instructed to collect MFP-QoL survey

data for all participants before the transition to the community, and again at one and two years after transition.

Methods and staff used to administer the survey vary by state. Grantees reported using one of four staff types to

administer the instrument and collect MFP-QoL data: (1) transition coordinators; (2) private contractors (such as

universities); (3) office-based staff (for example, people employed by the state Medicaid office); and (4) vol-

unteers. Use of transition coordinators is the most common approach, followed by use of office-based staff and

private contractors. Use of proxy respondents was permitted, and the survey was completed by proxy respondents

for 23 percent of pre-transition interviews and 16 percent of post-transition interviews.

For more information on this report, contact Carol Irvin at 617-301-8972 (or cirvin@mathematica-mpr.com).

Princeton, NJ Ann Arbor, MI Cambridge, MA Chicago, IL Oakland, CA Washington, DC

Visit the Mathematica website at: www.mathematica-mpr.com.

Mathematica

is a registered trademark of Mathematica Policy Research, Inc.