Challenges to community transitions through Money Follows the Person

Abstract

Objective

To examine the effects of transition challenges on the success and timeliness of transitions from institutions to community living for long‐stay participants in the Money Follows the Person (MFP) Rebalancing Demonstration and determine whether outcomes vary by age and disability.

Data Source

Secondary data on transition challenges for individuals enrolled in Connecticut's MFP program between December 2008 and December 2017.

Study Design

Challenges were analyzed for older adults, people with mental health disability, and people with physical disability. Bivariate and multivariate analyses investigated which transition challenges and selected demographic variables predict transition versus closure and length of transition period for each group.

Data Extraction Methods

The sample includes 3506 persons who attempted transition from institutions to community living and whose case concluded with transition or closure from 2015 to 2017.

Principal Findings

The association between most transition challenges and the ability of long‐stay institutional residents to return to the community, and to do so in a timely manner, varies significantly among older adults and younger persons with physical or mental health disabilities. For all groups, however, consumer engagement challenges predicted closure without transition (OR: 1.3‐3.9) and housing challenges predicted longer transition periods (84‐132 days). Length of institutional stay was associated with both outcomes for older adults and persons with physical disability. Other challenges, such as issues with services and supports, differed among the three groups on both outcomes.

Conclusions

Knowledge of the effects of transition challenges on success and timeliness of transition for each group allows program managers and health and service providers to focus resources on addressing the most serious challenges. Particular emphasis should be placed on consumer engagement and housing challenges, and on targeting persons for transition early in their institutional stay. Federal and state transition programs can benefit by individualizing supports for residents to yield successful outcomes.

What this study adds.

• While prior research has identified some factors that enhance or reduce the likelihood of institutional residents successfully transitioning back to community living, most studies involved short‐stay residents and did not distinguish among persons of different ages or disabilities. In addition, none has examined factors associated with the length of the transition period.

• This study examined challenges to the success of transitions and the length of the transition period in a large sample of long‐stay residents in one state's Money Follows the Person Rebalancing Demonstration. It also examined whether outcomes differed by target group, including older adults, persons with mental health disability, and persons with physical disability. The study found that most challenges vary significantly by group, and are associated with transition or length of transition period for at least one group. It recommends particular attention by transition program managers to consumer engagement and housing challenges, and to targeting persons for return to the community earlier in their institutional stay.

1. INTRODUCTION

State and federal policies regarding the provision of long‐term services and supports (LTSS) to older adults and people with disabilities are shifting. Recent efforts have focused on honoring individuals’ wishes to live in the community by “rebalancing” away from Medicaid's historical institutional bias toward home and community‐based settings.1, 2 Rebalancing strategies accelerated after the Supreme Court's 1999 Olmstead decision, which requires the provision of LTSS in the most integrated and least restrictive setting.3, 4 Some efforts emphasize prevention or delay of nursing home admissions by expanding home and community‐based services (HCBS) and diversion programs.1 Other efforts initiated transition programs that assist institutional residents to return to the community.5 Most state‐level transition programs began in the late 1990s.4, 6 One state‐level example, Minnesota's Return to Community Initiative, transitions non‐Medicaid nursing home residents to their communities, indirectly saving costs by avoiding or delaying Medicaid.7, 8, 9

The largest transition policy effort is the federal Money Follows the Person (MFP) Rebalancing Demonstration, created under the Deficit Reduction Act of 2005 and expanded by the 2010 Affordable Care Act, which has supported transitions to community living for Medicaid recipients in forty‐six states and the District of Columbia.10 MFP states select one or more “target groups” by age or disability and create protocols to identify participants and develop community HCBS care plans. All states identified adults age 65+ as a target group.

To effectively support institutionalized individuals’ return to the community, it is important to recognize and address challenges and determine whether the prevalence and impact of these challenges differ by age or disability. Some studies have examined factors that enhance or reduce the likelihood of returning to the community, but none have distinguished among individuals with different ages or disabilities. One salient factor is length of stay. Most individuals who leave a facility do so within 90 days of admission; transitioning after longer stays is much harder.11, 12 Studies have examined two subsets of the potential transition population: short‐stay residents (90 days or fewer) and long‐stay residents (91+ days).3, 13, 14 The most important resident‐level considerations associated with the probability of community transition are the resident's preference for community discharge and someone to support that preference,12, 13, 15 both of which are more prevalent in the early days after admission. Since many residents transition without assistance shortly after admission, Arling et al12 recommended that transition programs focus on individuals with 90‐ to 120‐day stays, when they are in danger of becoming long‐stay residents, but still want to leave and retain community ties. MFP requires a 90‐day institutional stay, allowing it to target individuals at that critical time.

A recent review concluded that persons who were younger, female, married, and received intense therapy were most likely to transition.16 Additional factors, particularly within 90 days of admission, include a supportive facility,12 availability and funding of HCBS,11 and supplemental community supports.4 Lack of affordable or accessible housing reduces the odds of transition,4, 17, 18, 19 although two recent studies did not confirm the relationship between housing and successful transition.7, 20 Other factors inhibiting transition include cognitive impairment,12, 13, 16 mental health diagnoses and substance use disorders,21, 22 and greater functional impairment or medical complexity.12, 13, 19

Fewer predictors of community discharge have been identified for long‐stay residents. Gassoumis et al13 found that only cancer or severe cognitive impairment reduced the odds of discharge after 90 days. Fries and James23 found the most predictive factors for return to the community after 90 days were desire to return, age under 84, quadriplegia/paraplegia, higher involvement in activities, and being in the least resource‐intense groups, plus the absence of schizophrenia or severely impaired cognition.23 Hass et al24 built and tested a model for Minnesota's MFP program to predict community discharge probabilities for nursing home residents at 90 days post‐admission, but did not include potentially modifiable factors such as lack of housing or caregiver availability. They agreed with Arling et al12 that preference for community discharge was the strongest indicator, and found additional predictive factors including age, health, and functional dependence.24

This dearth of information on transition predictors for long‐stay residents makes it challenging for transition planners to address issues faced by residents most in need of assistance. Because of its 90‐day stay requirement, MFP data on factors associated with successful transitions are particularly valuable. Moreover, since states identify MFP participants by target group, program data also detect challenges that differ by age or disability. For residents who use transition programs, no studies have examined factors associated with length of transition period. The national MFP evaluator, Mathematica Policy Research, did compare participating states on average length of transition period and found widely disparate results ranging from seven days to 13 months, due in part to varying state definitions.10 A closer examination of MFP data can identify factors that attenuate transitions and assist program managers in addressing transition challenges to achieve more timely results.

1.1. Research setting: Connecticut's MFP Demonstration

Data from Connecticut's MFP program are well‐suited to identify factors associated with the success and timeliness of transition for long‐stay residents. Connecticut was one of the earliest states to implement its MFP transition program, beginning in 2008. Upon referral, a transition team comprising the resident, family members, care manager, transition and housing coordinators, facility social worker, and community providers assesses the resident's support needs and creates a community care plan to address those needs and overcome identified challenges.25 During the transition process, Connecticut transition coordinators complete a standardized, cumulative checklist of transition challenges for each individual.26 Challenges are never overwritten or deleted, resulting in a cumulative list of all challenges recorded during the transition process, that is finalized at the time of transition or case closure without transition. Challenges are identified by the person, family members, providers, and MFP staff throughout the process until the person transitions or their case closes. A case may close without transition for several reasons, including death, participants changing their mind or refusing to cooperate with the transition process, or the person's conservator or legal guardian refusing to participate in the transition planning process. Fourteen challenge categories include physical health, mental health, financial, consumer engagement, services/supports, Medicaid waiver/HCBS issues, housing, legal/criminal, facility‐related, dementia/cognitive, family member/unpaid caregiver, other involved individuals, benefits/insurance, and substance abuse (Appendix S1).

Connecticut target groups include older adults aged 65+, persons with mental health disability, and persons with physical disability under age 65. (A small group of persons with developmental/intellectual disabilities are excluded from this analysis.) For all target groups from the inception of the MFP program, mean length of stay from facility admission to MFP referral has decreased steadily over the years 2009‐2017 (Table S1), reflecting both the unavailability of MFP for long‐stay residents prior to 2009, and increasing program outreach to residents earlier in their institutional stay. In the study population, described more fully below in Methods, the percent of each target group with each challenge category varies significantly in most cases (Figure 1), underlining the importance of identifying predictors of transition success and timeliness separately for older adults and younger individuals with different types of disabilities.

Figure 1.

Transition challenges by group. [Color figure can be viewed at http://wileyonlinelibrary.com]

Notes: ***P‐value <.001.

¹By definition, everyone in the mental health disability group has a mental health challenge and everyone in the physical diasbility group has a physical health challenge; therefore, these challenges were excluded from analyses for the respective groups

1.2. Research questions

Which transition challenges predict the likelihood of transition versus closure and length of transition period, and do predictors differ by age and disability? We hypothesize that (a) each challenge type predicts lower odds of transition; (b) for persons who transition, each challenge type predicts longer transition period; and (c) the effects of transition challenges vary by target group.

2. METHODS

The UConn Health institutional review board approved the study.

2.1. Data source

Data come from transition challenge checklists for individuals enrolled in Connecticut's MFP program between December 2008 and December 2017. Transition coordinators are autoreminded monthly to add any new challenges to the challenge checklist and are required to finalize it before transitioning or closing a case in the system; therefore, there are virtually no missing data. All challenges identified at any point in the process stay on the checklist for a final cumulative list of all challenges experienced.

2.2. Population studied

The study population includes 3506 individuals: older adults aged 65+ (n = 1806), persons with mental health disability (n = 192), and persons with physical disability under age 65 (n = 1508). Participants included in the analyses either transitioned from an institution to the community or closed without transitioning during 2015, 2016, or 2017.

2.3. Measures

The two outcome measures were community transition versus closure between 2015 and 2017 and, for persons who transitioned, length of transition period in days. Length of transition period was measured from the date of MFP referral to transition date. There were no censored cases; every case reached its final disposition. The primary independent variables of interest were the 14 transition challenge categories. Mental health challenges for persons with mental health disability and physical health challenges for persons with physical disability were excluded from analyses, by definition. Independent variables associated in the literature with community transition and demographic factors were included as follows: age at referral, length of stay in the facility from admission to MFP referral date, gender, race, and Hispanic origin.

2.4. Analysis

Chi‐square and independent‐samples t tests identified significant bivariate relationships between the independent variables and the two outcomes of interest: transition versus closure and length of transition period, in each of the three groups. Next, logistic regression models identified significant predictors of closure without transition for each group, and robust regression models identified predictors of length of transition period for each target group. The models included all challenges with a significant bivariate relationship for at least one target group, as well as age, length of stay, gender, race, and Hispanic origin. Substance abuse was not significant bivariately for any group for either outcome and was not included in multivariate analysis. Variance inflation factor (VIF) statistics confirmed the absence of issues of multicollinearity among all independent factors; all VIF values fell between 1 and 2, well below the accepted cutoff of 5. Robust regression methods were used to address the right skewness in the response variable. These methods provide stable results in the presence of outliers. The ROBUSTREG procedure in SAS was utilized with Huber M estimation which is commonly used when the outliers are primarily in the response direction. All other analyses were conducted in SPSS, version 25.

3. RESULTS

Table 1 compares baseline characteristics of MFP participants at the time of MFP referral by group and differences among the groups on the two dependent variables: transition versus closure and mean length of transition period. It also presents mean length of period before closure for persons who closed without transition. The target groups differed significantly by age, gender, race, Hispanic origin, and mean length of stay from facility admittance to MFP referral date. Older adults had a higher mean age at referral, the highest percentage of females, and the longest mean length of stay. Persons with physical disability were the most racially and ethnically diverse. Older adults were least likely to transition, but had the shortest average transition period, conditional on transitioning.

Table 1.

Characteristics of MFP participants (N = 3506)

	Older adult (n = 1806)	Mental health disability (n = 192)	Physical disability (n = 1508)
Baseline characteristics
Age (at MFP referral)***
Mean age	76	56	51
Minimum age	59	19	0
Maximum age	102	78	64
Gender, %***
Female	60	53	39
Male	40	47	61
Race, %***
White	79	83	71
Non‐white	21	17	29
Hispanic origin, %*	11	12	14
Mean length of stay (in months, facility admit to MFP referral date)***	9.0	6.9	5.6
Transition variables
Transitioned (percent)***	49	54	60
Closed without transition (percent)	51	46	40
Mean (SD) length of transition period (for those who transitioned, number of days from MFP referral date to transition)***	N = 881 242 (244)	N = 104 270 (167)	N = 908 317 (277)
Median (interquartile range) for length of transition period	166 (209)	232 (192)	236 (253)
Mean (SD) length of period before closure (for those who closed, number of days from MFP referral date to recommended closure date)***	N = 925 429 (332)	N = 88 451 (415)	N = 600 462 (400)

^* P < .05.

^** P < .01.

^*** P < .001.

A comparison of the prevalence of each challenge by target group is displayed in Figure 1. All challenges except mental health and services and supports varied significantly by group. Since everyone with a mental health disability had a mental health challenge, and everyone with a physical disability had a physical health challenge, these challenges were excluded from analysis for the respective groups. Housing was the most prevalent challenge for persons with mental health (86 percent) or physical disability (87 percent). Physical health challenges were the most prevalent for older adults (76 percent), with housing a close second (72 percent). Physical health challenges were the second most prevalent challenge for persons with mental health disability (63 percent). About half of all groups had a services and supports challenge; all other challenges were recorded for less than 50 percent of each group. The least prevalent challenge overall was other involved individuals, recorded for 10 percent or less of each group. Older adults were far less likely to experience financial, substance abuse, or facility‐related challenges than persons in the other two groups.

On average, older adults had 3.5 challenges, people with mental health disability had 4.3 challenges, and people with physical disability had 4.5 challenges. Individuals who closed without transition had mean number of challenges of 4.2 (older adult), 5.1 (mental health disability), and 4.6 (physical disability).

3.1. Factors associated with community transition

For each target group, bivariate analyses compared the percent of people with and without each challenge type who transitioned to the community (Table S2). Next, logistic regression models identified predictors of closure without transition for each group (Table 2). Older adults with challenges related to housing or family members/unpaid caregivers were about twice as likely to close. Older adults were also more likely to close if they had physical health, consumer engagement, or dementia/cognitive challenges. However, older adults with facility‐related or services/supports challenges were more likely to transition. Longer length of stay was associated with greater likelihood of closing, while non‐white and Hispanic older adults were more likely to transition.

Table 2.

Predictors of closure without transition by target population (logistic regressions)

Independent variable	Older adult (n = 1769)		Mental health disability (n = 185)		Physical disability (n = 1482)
	Odds ratio	95% CI	Odds ratio	95% CI	Odds ratio	95% CI
Physical health challenge	1.71***	1.33, 2.20	0.89	0.38, 2.07	NA
Mental health challenge	1.08	0.87, 1.35	NA		0.96	0.76, 1.21
Financial challenge	0.90	0.63, 1.29	0.32*	0.11, 0.92	0.68**	0.53, 0.88
Consumer engagement challenge	1.31*	1.05, 1.65	3.89**	1.71, 8.83	1.31*	1.02, 1.69
Services and supports challenge	0.55***	0.44, 0.69	0.64	0.26, 1.57	0.92	0.72, 1.17
Waiver/HCBS challenge	0.93	0.70, 1.24	6.38***	2.41, 16.91	1.12	0.85, 1.47
Housing challenge	1.96***	1.55, 2.48	1.49	0.50, 4.43	0.98	0.69, 1.37
Legal/criminal challenge	1.13	0.87, 1.48	0.79	0.33, 1.86	1.27	1.00, 1.62
Facility‐related challenge	0.66*	0.47, 0.91	1.75	0.66, 4.65	0.68**	0.52, 0.91
Dementia/cognitive issue challenge	1.52***	1.21, 1.92	2.13	0.77, 5.88	1.30	0.94, 1.80
Family member/unpaid caregiver challenge	2.05***	1.61, 2.62	1.05	0.39, 2.86	1.46**	1.10, 1.93
Other involved individuals challenge	1.03	0.63, 1.69	5.36*	1.20, 24.04	1.55	0.78, 3.06
Benefits or insurance challenge	1.20	0.95, 1.53	1.20	0.48, 3.00	0.75*	0.59, 0.96
Age at referral	1.06	1.00, 1.12	1.01	0.81, 1.26	1.16***	1.10, 1.23
Length of stay (months)	1.02***	1.01, 1.02	1.02	0.98, 1.07	1.01*	1.00, 1.02
Male	1.15	0.94, 1.42	0.54	0.24, 1.20	1.05	0.84, 1.32
Non‐white	0.94**	0.90, 0.98	0.89	0.74, 1.06	0.96	0.93, 1.00
Hispanic	0.54***	0.39, 0.75	4.34*	1.15, 16.40	1.16	0.85, 1.59

^* P < .05.

^** P < .01.

^*** P < .001.

People with mental health disability were considerably more likely to close without transitioning if they experienced challenges with consumer engagement (OR = 3.89), waiver/HCBS services (OR = 6.38), or other involved individuals (OR = 5.36), but 68 percent less likely to close if they had financial challenges. Those of Hispanic origin were over four times more likely to close. Two factors were independently related to greater likelihood of closing without transition in the physical disability group: consumer engagement and family/unpaid caregiver challenges. Three factors associated with greater likelihood of transition include financial, facility‐related, and benefits/insurance challenges. Higher age at referral and longer length of stay were associated with higher odds of closing.

3.2. Factors associated with length of transition period

For persons in each group who transitioned to the community, we explored factors associated with length of transition period (see Table S3 for bivariate results). Table 3 displays robust regression models for length of transition period for all groups. Financial and facility‐related challenges as well as gender and Hispanic origin have no independent effect on length of transition period for any group. Housing challenges are strong predictors of longer transition periods for all groups, adding between 86 and 132 days to the typical transition period, as are consumer engagement challenges for the mental health disability group (123 days) and physical disability group (34 days). Additional challenges predicting longer transition periods for older adults include legal/criminal, benefits/insurance, and family member/unpaid caregiver. For persons with physical disability, predictors of longer transition periods also include dementia/cognitive issues, waiver/HCBS issues, and services/supports challenges. For the older adult and physical disability groups, longer length of stay and non‐white race also predict longer transition periods, while higher age at referral predicts shorter transition periods for older adults. Persons with physical disability and a benefits/insurance challenge experienced shorter transition periods. The models explained about one‐fifth of the variance in length of transition period for all groups.

Table 3.

Predictors of length of transition period, in days, by target population (robust regression)

Independent variable	Older adult (n = 868)		Mental health disability (n = 102)		Physical disability (n = 897)
	B	SE	B	SE	B	SE
Financial challenge	28.65	15.76	0.93	37.41	–0.72	13.08
Consumer engagement challenge	3.71	9.34	122.55**	38.47	33.77*	13.36
Services and supports challenge	–9.50	8.74	4.87	37.02	34.45**	12.64
Waiver/HCBS challenge	1.04	12.41	44.49	57.52	54.62***	14.99
Housing challenge	86.34***	9.21	111.58*	46.80	132.20***	19.17
Legal/criminal challenge	81.16***	12.01	–1.87	34.65	21.86	12.62
Facility‐related challenge	11.72	13.09	–22.74	41.55	–12.04	13.97
Dementia/cognitive issue challenge	7.36	9.92	–56.01	52.97	57.59**	18.74
Family Member/unpaid caregiver challenge	37.00***	11.14	–56.93	45.15	7.13	15.73
Benefits or insurance challenge	61.35***	10.28	–44.95	38.10	–37.80**	12.60
Age at referral	–10.64***	2.47	4.36	10.77	2.20	2.56
Length of stay (months)	3.36***	0.39	4.71	3.39	10.05***	0.53
Male	0.30	8.59	−18.90	32.60	–9.71	11.93
Non‐white	3.23*	1.64	4.40	7.06	4.88*	2.07
Hispanic	8.04	12.83	115.03	62.56	16.97	17.02
Constant	245.16	41.23	74.38	138.09	46.59	35.44
R 2	0.21		0.22		0.19

^* P < .05.

^** P < .01.

^*** P < .001.

4. DISCUSSION

In the decade since Nishita et al15 published their findings, concluding that far more nursing home residents than previously estimated had the preference to transition to the community and the belief in their ability to do so, the MFP program has made such transitions a reality for thousands of people nationwide, fulfilling program goals of respecting choice and increasing engagement in community life.10 Typical survey comments from individuals who transitioned through Connecticut MFP, including “I loved getting out of there” and “I feel very fortunate to not be in the nursing home anymore,” testify to the important person‐level effects of program success. Yet many who attempt transition still fail to do so, or endure a lengthy transition process after a lengthy institutional stay. Knowledge of factors that prevent or lengthen transitions for various groups can help transition programs target challenges more precisely and concentrate their limited resources on the most prevalent and addressable barriers to transition.

Our research extends previous literature by quantifying the prevalence of transition challenges by target group, identifying additional transition predictors for long‐stay residents, specifying factors salient to older adults compared to younger adults with physical or mental health disability, and identifying transition challenges that increase length of transition period. Our hypothesis that the effects of transition challenges vary by target group was confirmed by our findings, but our hypothesis that each challenge type predicts lower odds of transition and longer transition periods was confirmed only in part. In contrast to previous literature,21, 22 for example, substance use was not a significant predictor of either outcome for any group.

Confirming previous literature, older adults were significantly more likely to close without transitioning if they had physical health or housing challenges or dementia/cognitive issues, but none of these challenges were associated with likelihood of transition for the other groups. Longer length of stay, as also noted in the literature, was associated with closure without transition for older adults and persons with physical disability. Our additional finding that consumer engagement challenges, for all three groups, and issues with family members or unpaid caregivers, for older adults and persons with physical disability, are also associated with greater odds of closing supports Arling et al's12 conclusion that preference for returning to the community and someone to support that preference are critical.

Persons with mental health disability had few but strong predictors of not transitioning. In addition to challenges with consumer engagement, only Medicaid waivers/HCBS issues, other involved individuals, and Hispanic origin were associated with closure before transition. Although not prevalent, waiver and other involved individual challenges are significant only for this group and should be a focus for programs aiming to increase transitions for persons with mental health disability. Some mental health waivers have highly specific eligibility criteria or may not adequately meet the consumer's needs, and mental health issues can lead to lack of motivation or unrealistic expectations regarding community supports. Other individuals involved with a consumer who may not support transition for this group include their physicians and mental health providers. They should be integrated into the transition process allowing for a more informed and better educated opinion regarding transition for an individual.

Non‐white and Hispanic older adults are more likely to transition, possibly reflecting cultural differences in the propensity of family members to care for older relatives. Research on caregiving among various non‐white ethnic groups notes cultural themes of familial reciprocity, filial piety, and care expectations,27 which may account for greater willingness to care for older relatives in the community, and the increased likelihood of transition for non‐white and Hispanic older adults. Some ethnic groups may also lack formal caregiving supports and the resources to pay privately for assistance27; for these families, MFP may be a means of addressing racial/ethnic disparities by providing access to supports to care for relatives at home. In contrast to older adults, persons with mental health disability of Hispanic origin are far more likely to close (OR = 4.34). Stigma of mental illness in the Hispanic community may create less willingness to support them post‐transition. Related research notes some cultural beliefs that associate mental illness with danger, lack of control, and ostracism.28, 29 The striking contrast between Hispanic older adults and persons with mental health disability warrants further research into possible systemic disparity.

Unexpectedly, four challenges were associated with increased odds of transitioning, including services/supports (for older adults), facility‐related challenges (for older adults and persons with physical disability), financial challenges (for persons with mental health or physical disability), and benefits/insurance (for persons with physical disability). Services/supports challenges, such as lack of transportation or home care services, become evident later in the transition process and may not be recorded for individuals who closed earlier for other reasons. Similarly, facility/staff resistance or lack of discharge planning may occur only for individuals close to leaving the facility. Both facility‐related and benefits/insurance challenges may reflect the fact that facilities may have systemic incentives to accelerate transition for residents who do not meet institutional level of care for Medicaid eligibility, leading to a higher transition success rate. Financial challenges such as lack of income and poor consumer credit also surface later in the transition process when seeking apartment leases. People who close earlier in the process may not encounter these issues.

Predictors of length of transition period are relevant only for individuals who successfully transitioned. Housing challenges, among the most prevalent, predicted longer transition periods for all groups, supporting previous research that accessible, affordable housing is a key barrier for persons leaving institutions. For older adults, family and unpaid caregiver challenges not only increased the odds of closing, but also significantly lengthened the transition period. Two challenges that did not predict closure for older adults (legal/criminal and benefits/insurance) were associated with significantly longer transition periods. While these issues may not prevent transition, they take longer to resolve, particularly Medicaid eligibility, probate court or legal representative issues, and locating missing documents. In contrast to previous literature,16, 23 higher age at referral predicted shorter transition periods for older adults. This unexpected finding may partially reflect older adults near the end of life transitioning quickly with MFP supports so they can die at home. The only additional predictor of longer transition periods for persons with mental health disability was consumer engagement.

Medicaid waiver/HCBS, services/supports challenges, and dementia/cognitive challenges did not predict closure but did predict longer transition periods for the physical disability group. This group may not meet eligibility for targeted waivers, and it may take longer to arrange Medicaid state plan benefits to meet their physical and cognitive support needs. Unlike older adults and persons with mental health disability, who normally have agencies to arrange support needs such as personal care assistants (PCAs) and home health aides, younger persons with physical disabilities most often self‐direct their care. Before they can transition, they or their families must find, vet, hire, and train a full complement of PCAs, which may account for their longer transition times. Higher involvement by waiver case managers for older adults and persons with mental health disability plays a large role in connecting those groups to agency services; case management or other assistance with quickly hiring PCAs could usefully be considered by transition programs to speed up transitions for the physical disability group.

Interestingly, benefits/insurance challenges predicted longer transition periods for older adults but shorter periods for the physical disability group. Older adults often experience these challenges when applying for community Medicaid, due to the extensive time needed to compile necessary paperwork and numerous financial entanglements which can complicate Medicaid eligibility. In contrast, facilities have incentives to move out younger adults who do not meet Medicaid nursing facility level of care. This challenge both predicts greater odds of transitioning for the physical disability group and relates to shorter transition periods as facilities move them out quickly.

Among the most important and actionable findings, consumer engagement challenges strongly predict closure for all groups and significantly extend transition periods for the mental health and physical disability groups. Engagement can be addressed by program personnel through motivational interviewing. While motivational interviewing is standard for many care managers, it is not standard for transition and housing coordinators, who would benefit from expanded training and practice in this discipline.

Housing challenges significantly predict closure only for older adults, but strongly relate to longer transition periods for all groups, adding over 4 months on average. Recent research that found housing not to be a barrier involved either short‐stay private pay residents,7 with higher socioeconomic status, or options counseling clients,20 who may not have lost their housing. For long‐stay Medicaid residents, housing remains a prominent challenge to timely transitions and should be a primary focus of transition program resources.

Our finding that length of stay from facility admission to MFP referral is associated with both higher odds of closure and longer transition periods for older adults and people with physical disability illustrates the importance of targeting institutional residents earlier in their stay for assistance with transitioning. While length of stay has steadily decreased over the nine years of Connecticut's MFP program (Table S1), it may still be too long, especially for older adults, to avoid the challenges that put transition at risk. Transition programs should aim to begin far earlier by informing institutional residents of their options within 90 days of admission. Some states have begun that process, including Connecticut's new My Care Options initiative, which will target people 45‐60 days after admission. Similarly, Minnesota's Return to Community initiative targets private pay nursing home residents at 60‐90 days post‐admission,7, 8, 9 and its MFP program targets Medicaid‐eligible participants prior to 90‐day eligibility,24 with promising results.

This study has some limitations, reflecting experiences from one state for persons with Medicaid. Despite protocol training for populating the checklist, recording challenges involves some judgment and may not be completely uniform. Moreover, the checklist is cumulative. Challenges that do not manifest until late in the process would not occur if the case closed earlier for unrelated reasons. Just reaching the point of assessing for these later challenges may indicate that people have already overcome an initial set of challenges and are therefore more likely to succeed. In addition, the mental health disability group has a much smaller number of cases than the other groups, which diminishes the power to detect significant effects. A larger sample of cases with mental health disabilities may identify additional factors associated with their transition success.

5. CONCLUSION

Study results demonstrate which predictors of transition and length of transition period prevent or delay transition for which participants. People working to help individuals move from institutions to the community, including transition program administrators or health and social service providers, may use these findings to focus scarce program resources on targeted supports for the most serious challenges, yielding successful outcomes for all target groups. While these challenges may differ somewhat by programs, resources aimed at increasing consumer engagement, expanding housing options, and targeting institutional residents early in their stay should yield improvements in both the rate of successful transition and length of the transition process.

CONFLICT OF INTEREST

No other disclosures.

Supporting information

 

 

Robison J, Shugrue N, Porter M, Baker K. Challenges to community transitions through Money Follows the Person. Health Serv Res. 2020;55:357–366. 10.1111/1475-6773.13267