IN-HOME-PD Caregivers: The Effects of a Combined Home Visit and Peer Mentoring Intervention for Caregivers of Homebound Individuals with Advanced Parkinson’s Disease

Abstract

Introduction:

Family caregivers of people with advanced Parkinson’s Disease (PD) are at high risk of caregiver strain, which independently predicts adverse patient outcomes. We tested the effects of one year of interdisciplinary, telehealth-enhanced home visits (IN-HOME-PD) with 16 weeks of peer mentoring on caregiver strain compared with usual care.

Methods:

We enrolled homebound people with advanced PD (PWPD) and their primary caregiver as IN-HOME-PD dyads. We trained experienced PD family caregivers as peer mentors. Dyads received four structured home visits focused on advanced symptom management, home safety, medications, and psychosocial needs. Starting at approximately four months, caregivers spoke weekly with a peer mentor for 16 weeks. We compared one-year change in caregiver strain (MCSI, range 0-72) with historical controls, analyzed intervention acceptability, and measured change in anxiety, depression, and self-efficacy.

Results:

Longitudinally, IN-HOME-PD caregiver strain was unchanged (n = 51, 23.34 (SD 9.43) vs. 24.32 (9.72), p = 0.51) while that of controls worsened slightly (n = 154, 16.45 (10.33) vs. 17.97 (10.88), p = 0.01). Retention in peer mentoring was 88.2%. Both mentors and mentees rated 100% of mentoring calls useful, with mean satisfaction of 91/100 and 90/100, respectively. There were no clinically significant improvements in anxiety, depression, or self-efficacy.

Conclusions:

Interdisciplinary telehealth-enhanced home visits combined with peer mentoring mitigated the worsening strain observed in caregivers of less advanced individuals. Mentoring was met with high satisfaction. Future caregiver-led peer mentoring interventions are warranted given the growing, unmet needs of PD family caregivers.

Trial registration:

NCT03189459

1. Introduction

The overwhelming majority of the 1.2 million people living with Parkinson’s disease (PWPD) in the US rely on unpaid family caregiving, at an estimated cost of US $6.5 billion in caregiver lost wages alone.^1,2 Parkinson’s disease (PD) caregiving is associated with significant physical, psychological, and financial burdens.^3,4 As PD progresses, quality of life declines and hospitalizations increase.^5-8 Worsening PD symptoms exacerbate caregiver strain, which further aggravates PD symptoms and independently increases PWPDs’ risk of hospitalization and institutionalization.^9-14 Caregiving is particularly burdensome in late-stage PD, which is marked by increasing dependence for activities of daily living. Such caregivers report anxiety, depression, pain, and higher caregiver burden, though little longitudinal data exists on late-stage PD caregivers.¹⁵ Hence, PD family caregivers are a population ripe for intervention and a critical inflection point for improving patient outcomes. The need for effective and pragmatic family caregiver interventions is urgent, yet no PD-specific caregiver interventions have demonstrated efficacy.^16,17

We previously piloted home visits for homebound, advanced PWPD, which appeared to uncouple the expected decline in quality of life from the downward trajectory of mobility and function.^18,19 Here, “homebound” individuals—based on the US Centers for Medicare and Medicaid Services definition—are those for whom leaving the home is a considerable and taxing effort due to their medical condition, and for whom leaving requires the assistance of another person, assistive devices (e.g., wheelchair), or assistive transportation.²⁰ Despite a labor-intensive, longitudinal approach with nursing, social work, and movement disorders expertise, caregiver strain worsened over one year. Without other caregiver data, it is unclear whether home visits were ineffective on caregiver strain or mitigated an otherwise steeper decline. We hypothesized that increasing needs and complexity paired with estrangement from outpatient medical care experienced by homebound PWPD^21,22 means that access to continued care through home visits may be necessary but insufficient for supporting such caregivers. In designing the next step in the home visit model, we sought to both: 1) enhance the model with an evidence-based, caregiver-focused intervention, and 2) compare longitudinal progression in caregiver strain among those receiving the combined intervention vs. caregivers whose PWPD are receiving usual care.

In other serious illnesses, peer mentoring interventions have benefited patients and caregivers. In the United Kingdom, caregivers of individuals with Alzheimer’s Disease and Alzheimer’s Disease related dementias were paired with peer mentors—former informal caregivers to a loved one with dementia—to bolster coping skills and social support.²³ Qualitative data supported significant benefits for both the caregiver mentees and mentors.^24,25 Subsequently, peer mentoring has improved outcomes in trauma survivors, individuals with kidney disease,^26-28 sedentary older adults, young adults with neurologic disorders,^29,30 and family caregivers of individuals with cancer.³¹ While the structure, duration, and populations vary, each presupposes that interacting with a peer who has navigated similar situations can motivate behavior change. Indeed, a recent survey among PD caregivers indicates both a desire among current caregivers for a peer mentor, and a willingness among former PD caregivers to serve such a role.³²

Therefore, to better address both PWPDs’ and caregivers’ needs, we studied one year of structured, telehealth-enhanced, interdisciplinary home visits for homebound PWPD-caregiver dyads in which 16 weeks of peer mentoring (“Share the Care”) occurred between months 4-8.^33-35 Home visits focused on symptom management, home safety, medications, and psychosocial needs of the dyads; peer mentoring focused on support and coaching by an experienced caregiver trained by the study team. We hypothesized that the combined intervention would improve caregiver strain at one year compared with PWPD-caregiver dyads in a national longitudinal registry. We also hypothesized that peer mentoring would be 1) feasible, measured by adoption and retention, 2) acceptable, measured by mentor and caregiver satisfaction, and 3) effective, measured by change in caregivers’ strain, anxiety, depression, and self-efficacy.

2. Methods

2.1. Study design

IN-HOME-PD was a nonrandomized controlled trial evaluating the feasibility and efficacy of four structured, interdisciplinary, telehealth-enhanced home visits for homebound PWPD-caregiver dyads over approximately one year. Given the risk of homebound individuals becoming lost to follow-up due to death or institutionalization²² and high dropout rates associated with waitlist control arms in PD trials,³⁶ a randomized design was rejected. Rather, we sought a historical control group with a large enough volume of longitudinally-followed PWPD-caregiver dyads with which to compare our study participants. Enrolled dyads were compared with historical controls drawn from the Parkinson’s Outcomes Project (POP) registry in which PWPD and their caregivers complete brief, annual assessments while receiving usual care at a Parkinson’s Foundation Center of Excellence. Given variability in healthcare systems, and definitions of and services available to homebound individuals in different countries, we limited our pool of potential control dyads to those cared for only at US Centers of Excellence. Using the baseline characteristics of participants in our home visit pilot study¹⁸ and the >10,000 PWPD in the POP at the time of study design, we projected the ability to match over 90% of homebound participants in the planned study with at least two POP participants based on sex, age (± ten years), and Hoehn & Yahr (HY) stage. The caregivers of the matched POP participants would then serve as the control group for caregiver comparisons in the absence of other PD caregiver-specific registries.

Share the Care refers to the 16-week nested PD caregiver peer mentoring intervention occurring between IN-HOME-PD visits 2-3 (approximately months 4-8 of study participation). While our power analysis was designed to compare the full one year intervention among IN-HOME-PD caregivers compared with one year of usual care in POP caregivers, we also sought to explore the within-group change in IN-HOME-PD caregiver strain over one year vs. interval change with home visits alone, visits plus mentoring, and visits post-mentoring (Figure 1). No similar peer mentoring interventions exist at any of the POP enrollment sites; however, all caregivers could participate in support groups and other activities. The protocols for IN-HOME-PD and Share the Care, and the PWPD feasibility and quality of life outcomes have been previously published.^33-35

Figure 1:

Study design, activities, and measures for IN-HOME-PD Caregivers

2.2. Participants, recruitment, and control population

Table 1 describes the participant categories (IN-HOME-PD PWPD and caregivers/mentees, and historical PWPD-caregiver dyads), recruitment methods, and inclusion and exclusion criteria. The caregiver member of each IN-HOME-PD dyad was the mentee in Share the Care. Once the 65 IN-HOME-PD dyads were recruited and enrolled, the sex, age, and HY stage of each PWPD was sent to the POP study team for selection of matched controls. The POP does not assess for homebound status and thus we could not match based on this definition, which would have been preferable. One-to-one matching of IN-HOME-PD and POP PWPD proved impossible: despite >13,000 enrolled POP participants, the number of participants staged as HY ≥4 was extremely limited, and among those, individuals with longitudinal follow-up (defined as at least one additional visit between 9 to 18 months after the index visit with HY stage 4) were exceedingly rare. Our pilot matching algorithm had assumed that the majority of PWPD HY ≥4 would be retained for at least two consecutive visits, yet this was not the case. Multiple statistical methods to account for this—including propensity score matching—were attempted by the study biostatistician (BO) and failed. Thus, in the absence of one-to-one matching, we requested all US POP PWPD participants with at least two consecutive visits with a Hoehn & Yahr (HY) stage of 3 or higher at the index visit.³⁷ This pool of 319 individuals, heavily skewed towards Hoehn & Yahr 3, formed the control group for between-group comparisons of PWPD outcomes³⁴ and the caregivers of each dyad—when present—formed the caregiver control group for between-group comparisons described herein.

Table 1:

Study participants, recruitment methods, and inclusion/exclusion criteria

	Intervention Group		Control Group
	IN-HOME-PD Person with Parkinson’s Disease	IN-HOME-PD Caregiver/Share the Care Mentee	POP Person with Parkinson’s Disease Controls	POP Caregiver Controls
Definition	Individual with PD* meeting Centers for Medicare & Medicaid Services homebound criteria, seen at least once at Rush University	Unpaid, informal caregiver of homebound person with advanced PD	Individual with PD* receiving care at one of 16 US Parkinson’s Foundation Centers of Excellence and enrolled in the Parkinson’s Outcomes Project registry	Informal caregiver of person with PD
Recruitment	Recruited as dyad based on screening of electronic medical record and direct referrals from Rush University movement disorders specialists		POP participants recruited from individuals receiving care for PD at Parkinson’s Foundation Centers of Excellence by site study coordinators; retrospective POP data queried for this study based on care recipient meeting study-specific inclusion criteria below
Inclusion Criteria	English-speaking ≥40 years old Hoehn & Yahr ≥3 at most recent outpatient visit Informal caregiver willing to serve as study partner Community-dwelling Meets CMS homebound criteria1 ≥1 of the following evidence-based risk factors for loss to follow-up, hospitalization, or institutionalization, as determined by referring neurologist: Motor or cognitive fluctuations Multi-morbidity Medication mismanagement Cognitive impairment Depression and/or anxiety Suspected elder abuse Suspected caregiver burnout Recent history of increased falls at home ≥2 canceled or no-show appointments in past 12 months neurologist’s concern for high risk of hospitalization or institutionalization	English-speaking ≥30 years old Living in Chicago Cohabiting with patient or spending 20+ hours per week engaged in care-related tasks Working telephone number for study team & peer mentor to contact caregiver Capacity to consent	English-speaking Hoehn & Yahr stage ≥3 at index visit At least one subsequent visit ≥9 months after index visit	English-speaking Willing and able to complete annual POP questionnaires
Exclusion Criteria	Non-English-speaking Enrolled in hospice Assisted living or nursing home placement at the time of screening Atypical, vascular, or drug-induced parkinsonism Lack of informal caregiver	<1 year life expectancy Severe psychiatric disorder Primarily non-English speaking	Care recipients unable or unwilling to give written informed consent POP participant enrolled at non-US site	Care recipients unable or unwilling to give written informed consent

^*According to treating movement disorders neurologist per UK Brain Bank criteria²

Peer mentors were current or former caregivers for PWPD or related disorders, aged 30 or older, who had cared for their loved one for at least two years, participated in support groups or similar educational or social activities as evidence of dedication to informal peer support, primarily English-speaking, and willing to commit to a daylong in-person mentor training session and one to two sequential 16-week blocks of peer mentoring. Current or former caregivers meeting peer mentor eligibility criteria were referred by Rush neurologists or identified through a search of the electronic medical record for caregivers of individuals with advanced PD (seen within three years at the Rush Center of Excellence), with additional referrals of outstanding support group leaders meeting eligibility criteria from a related stakeholder foundation.

2.3. Screening and informed consent

The Rush University Medical Center Institutional Review Board approved this study in October 2017. The team spoke directly with potential participants to discuss the study and confirm eligibility and interest. Caregiver mentees and peer mentors completed written informed consent, at home visit 1 or mentor training, respectively. POP controls consented to registry participation and their data was queried retrospectively.

2.4. Interventions

2.4.1. IN-HOME-PD Home Visits

Briefly, IN-HOME-PD dyads received four protocol-driven home visits from a nurse and a social worker, accompanied by telehealth connection with a movement disorders specialist, over one year. Caregiver-focused portions of visits included an in-depth psychosocial needs assessment involving socioeconomic background, support network, and community and healthcare services received and needed. The nurse conducted a detailed medication reconciliation and home safety assessment, identifying medication errors, non-adherence, expired drugs, or suboptimal organization, and fall, fire, and other home safety risks. The team provided extensive caregiver and PWPD education and counseling on identified issues and changes to the treatment plan. This included a short handbook focused on common symptoms and scenarios prompting urgent or emergent care, and approaches to recognition and management (e.g., urinary tract infections). The social worker offered referrals to community services available through state and local jurisdictions, including Meals on Wheels^®, friendly visitors, lending closets for durable medical equipment, and counseling services. The nurse called the caregiver 2-4 weeks after each home visit to manage any new questions, concerns, or unmet needs. Caregivers could contact the study team between visits by phone or the electronic health portal.

2.4.2. Share the Care Peer Mentoring

2.4.2.a. Peer mentor training:

Potential peer mentors attended one five-hour, in-person mentor training session comprised of 1) didactics on the stages, symptoms, and management of advanced PD, created and led by the movement disorders neurologist (JEF), to ensure all mentors had a similar baseline level of familiarity regardless of their own caregiving experience; 2) interactive session on a mentor’s role and skills, led by the study social worker (JL); and 3) review of the content and use of the Share the Care mentoring handbook,³⁵ led by the study social worker and coordinators (JL, ECK). Mentors could contact their mentees by phone call or videoconferencing apps using a study-provided iPad or their personal devices.

2.4.2.b. Mentor matching:

Once trained, mentors were matched with one or two mentees sequentially over one year. Each pairing lasted 16 weeks. When possible, matches were based on relationship to the care recipient, age, or geographic location to facilitate potential in-person meetings.

2.4.2.c. Peer mentoring:

Mentors were asked to speak with their mentees for 30 minutes weekly by phone, videoconference, or in person. Mentors could focus on topics of mentee concern and refer to the Share the Care handbook to guide discussions. Participants maintained brief study diaries documenting the date, duration, utility of each conversation, and any concerns. Mentors were invited to quarterly conference calls to share their experiences with each other and the study team. Additional details about the Share the Care mentor training, handbook, and mentoring process are described elsewhere.³⁵

2.5. Measures

Figure 1 depicts the IN-HOME-PD and Share the Care activities and assessments for caregiver and mentor participants at each study timepoint. Caregivers completed demographic questionnaires at home visit 1 (baseline) and primary and secondary outcome measures at each of the four home visits, including baseline, 4 months, 8 months, and approximately 12 months. Peer mentoring occurred between home visits 2-3 (months 4-8) and satisfaction with mentoring was assessed at home visit 3. Mentors completed demographic questionnaires and secondary outcome measures at mentor training (0-6 months before their first mentoring relationship began), with measures repeated and a satisfaction measure added after 16 weeks of mentoring. Caregivers and mentors completed biweekly study diaries during weeks 2-16 of mentoring. Demographics included age, sex, race, ethnicity, highest level of education, marital status, relationship to PWPD, and caregiving duration. For POP caregivers, only relationship to the PWPD and caregiving duration were available.

We collected all self-reported caregiver measures at home visits 1-4. Our primary efficacy measure for IN-HOME-PD caregivers was caregiver strain—both overall and dimension-specific strain—as measured by the Multidimensional Caregiver Strain Index (MCSI).^38,39 MCSI is an 18-item self-report measure validated in PD caregivers and used at annual POP visits. Items are rated on a 4-point Likert scale, from 0 (never) to 4 (a great deal), total score 0-72, and categorized as mild (<20), moderate (20-29), and severe (≥30).⁵ MCSI dimensions include physical, financial, interpersonal, and demanding/manipulative strain, and time and social constraints.

Share the Care feasibility measures included adoption and retention rates and frequency and duration of mentor-mentee calls. Satisfaction was assessed using the Client Satisfaction Inventory Short Form (CSI-SF),⁴⁰ a 9-item, self-report measure, with items rated on a 7-point Likert scale from 1 (none of the time) to 7 (all of the time). CSI-SF scores range from 0-100, with higher scores indicating greater satisfaction. Mentees and mentors also indicated whether mentoring was useful (yes/no), they were “getting along with” their matched mentee or mentor (yes/no), and if any problems arose.

Secondary outcomes included mental health using the Hospital Anxiety and Depression Scale (HADS)⁴¹ and self-efficacy using the Caregiver Self-Efficacy Scale (CSES).⁴² The HADS is a 7-item, self-report, validated measure of anxiety and depression; scores ≥8 on either subscale indicate probable symptoms. The CSES is a 9-item, self-report measure of caregivers’ belief in their ability to succeed in specific situations or accomplish a task. CSES items are rated on a 10-point Likert scale from 1 (not at all certain) to 10 (very certain), with subscales for symptom management and community support.

2.6. Statistical Analyses

Sample size calculations were based on PWPD outcomes for the overall IN-HOME-PD study.³³ For caregiver-related outcomes, assuming a change of ten points in the MCSI from 24-34 (SD 8) over one year in controls based on our original home visit data,¹⁹ and alpha of 0.05, 50 caregivers afforded 99.9% power to detect a significant change in caregiver strain.

Baseline characteristics of each group are summarized as frequencies and percentages for categorical variables and mean/SD or median/interquartile range for continuous variables, as appropriate. Using paired and two-sample t-tests, we analyzed our primary outcomes: one-year change in overall and dimension-specific MCSI scores within IN-HOME-PD caregivers, and the difference in one-year change between IN-HOME-PD and POP caregivers. Peer mentoring feasibility metrics are summarized with descriptive statistics; if discrepant values were reported by the matched mentor and mentee, the average of the two reported values was used.

Among IN-HOME-PD caregivers only, our secondary outcomes included interval change in strain, anxiety, depression, and self-efficacy between each home visit with repeated measures ANOVA to account for within-subject variability. Additionally, we evaluated the percentage of caregivers in each MCSI strain category (mild, moderate, and severe) at each visit. Finally, we used paired t-tests to explore whether the process of peer mentoring improved peer mentors’ depression, anxiety, or self-efficacy.

3. Results

3.1. Descriptive analyses.

Sixty-five PWPD-caregiver dyads enrolled in IN-HOME-PD, of whom, 45 completed peer mentoring. Of the 65 enrolled caregivers, 51 completed all home visits (78.5% retention rate). Among the 14 caregivers who withdrew, nine were due to patient death. As illustrated in the CONSORT diagram in Figure 2, six caregivers completed all home visits but declined mentoring citing lack of time or interest.

Figure 2:

CONSORT diagram of peer mentor and caregiver mentee recruitment

Table 2 lists the baseline characteristics of the PWPD and caregivers in both the intervention and control groups. Among more than 13,000 PWPD in the POP registry, only 319 had two consecutive visits for which HY ≥3 at the index visit. Thus, the 65 IN-HOME-PD PWPD were compared with 319 POP controls, however this was a notably non-equivalent comparison group as POP was heavily skewed towards younger and less severely affected individuals.³⁴ IN-HOME-PD PWPD were significantly older, more racially diverse, and had more severe PD (78.5% vs. 15% HY ≥4), worse quality of life, and poorer cognition at baseline compared with POP controls (p ≤ 0.001 for each characteristic). Among POP controls, 290 had a caregiver for at least one visit and 154 had a caregiver providing complete information for two consecutive visits. Nearly all (99.7%) POP caregivers identified as the PWPD spouse/partner, with a longer median caregiving duration (8 years, IQR 5-11) than IN-HOME-PD caregivers (5 years, IQR 3-7). IN-HOME-PD caregivers had higher baseline strain than POP caregivers (MCSI total score: 22.5 (SD 9.5) vs. 16.5 (10.3), p <0.001).

Table 2:

Baseline Characteristics of IN-HOME-PD and Parkinson’s Outcomes Project People with Parkinson’s Disease and Caregivers

Characteristic		Intervention		Control
		IN-HOME PD Caregivers (n=65)	IN-HOME-PD People with PD (n=65)	POP People with PD (n=319)	POP Caregivers (n=290)
Mean age, years (SD) †		64.1 (11.5)	78.9 (7.6)	70.1 (7.8)	≠
Female, n (%)		51 (78.5%)	21 (32.3%)	121 (37.9%)	≠
Race, n (%) †	White	47 (72.3%)	47 (72.3%)	307 (96.2%)	≠
	Black	10 (15.4%)	9 (13.8%)	1 (0.3%)	≠
	Asian	6 (9.2%)	7 (10.7%)	7 (2.2%)	≠
	Native Hawaiian/Pacific Islander	0 (0%)	0(0)	1 (0.3%)	≠
	More than one race	0 (0%)	0 (0)	0 (0)	≠
	Unknown/not reported	2 (3.1%)	2 (3.1%)	3 (0.9%)	≠
Ethnicity, n (%)	Hispanic/Latinx	4 (6.2%)	4 (6.2%)	11 (3.4%)	≠
Education, n (%) †	Less than high school graduate	0 (0%)	9 (13.8%)	6 (1.9%)	≠
	High school graduate	10 (15.4%)	9 (13.8%)	53 (17.2%)	≠
	Some college/trade school	23 (35.4%)	12 (18.5%)	78 (25.2%)	≠
	Bachelor's degree	6 (9.2%)	10 (15.4%)	82 (26.5%)	≠
	Graduate degree	25 (38.5%)	25 (38.5%)	90 (29.1%)	≠
Marital status, n (%) †	Single	5 (7.7%)	3 (4.6%)	19 (6.0%)	≠
	Married/living with domestic partner	54 (83.1%)	44 (67.7%)	268 (84%)	≠
	Widowed	3 (4.6%)	14 (21.5%)	17 (5.3%)	≠
	Divorced/separated	4 (6.2%)	(6.2%)	15 (4.7%)	≠
Median PD duration (IQR) *		n/a	15 (10)	11 (7)	n/a
Hoehn & Yahr Stage, n (%) †	3	n/a	14 (21.5%)	271 (85%)	n/a
	4	n/a	41 (63.1%)	40 (12.5%)	n/a
	5	n/a	10 (15.4%)	8 (2.5%)	n/a
Abbreviated Montreal Cognitive	Immediate 5-item recall†	n/a	3.5 (1.4)	4.4 (0.9)	n/a
	Delayed 5-item recall†	n/a	1.9 (1.7)	3.5 (1.4)	n/a
Assessment Items, mean (SD)	Verbal fluency†	n/a	10.9 (6.1)	17.2 (6.3)	n/a
Baseline mean quality of life (PDQ-39 total score (SD)) †		n/a	38.0 (14.1)	29.5 (14.6)	n/a
Relationship to care recipient, n (%) †	Spouse/partner/significant other	39 (60%)	n/a	n/a	289 (99.7%)
	Adult child	19 (29.2%)	n/a	n/a	0 (0)
	Other family member/friend/neighbor/ home health aide	8 (12.3%)	n/a	n/a	1 (0.3%)
Median years of caregiving, (IQR)		5 (3-7)	n/a	n/a	8 (5-11)
Baseline mean caregiver strain (MCSI total score (SD)) †		22.5 (9.5)	n/a	n/a	16.5 (10.3)

^≠:variable not gathered on POP caregivers

^*:IN-HOME-PD vs. POP control two-tailed t-test p <0.05

^†:IN-HOME-PD vs. POP control two-tailed t-test p ≤ 0.001

IN-HOME-PD: Interdisciplinary Home Visits for Parkinson’s Disease; IQR: Interquartile range; MCSI: Multidimensional Caregiver Strain Index, higher score represents more severe strain; PDQ-39: Parkinson’s Disease Questionnaire, higher score represents worse quality of life; POP: Parkinson’s Outcomes Project; SD: Standard Deviation

3.2. Primary outcomes: Change in overall and dimension-specific caregiver strain between baseline and one year

Table 3 presents overall and dimension-specific strain at baseline and one year for both IN-HOME-PD and POP caregivers. At both time points, the intervention group fell in the moderate strain category and the control group, mild. Longitudinally, however, intervention caregivers’ strain was unchanged between baseline and one year (23.34 (9.43) vs. 24.32 (9.72), p = 0.51) while controls worsened slightly (16.45 (10.33) vs. 17.97 (10.88), p = 0.01) (Table 3). There were no changes in any specific dimensions of strain among the interventional group over one year (p = 0.33-0.96), and while physical strain and time constraint dimensions worsened among POP caregivers over one year (p = 0.03 and 0.02), between-group comparisons were not significant nor were the within-group changes clinically significant.

Table 3:

Comparison of overall and dimension-specific caregiver strain between IN-HOME PD and POP caregivers

	IN-HOME-PD Caregivers				POP Caregivers
	N	Baseline	1 year	p1	N	Baseline	1 year	p1	p2
Multidimensional Caregiver Strain Index, mean (SD)
Total caregiver strain	47	23.34 (9.43)	24.32 (9.72)	0.51	154	16.45 (10.33)	17.97 (10.88)	0.01*	0.73
Physical strain	49	4.27 (3.09)	4.69 (2.75)	0.33	156	2.76 (2.54)	3.13 (2.96)	0.03*	0.90
Social constraints	48	7.83 (3.41)	8.13 (3.69)	0.60	156	5.93 (3.86)	6.23 (3.86)	0.16	0.99
Financial strain	49	1.41 (1.64)	1.51 (1.84)	0.68	154	0.62 (1.07)	0.71 (1.08)	0.28	0.97
Time constraints	49	4.61 (1.82)	4.55 (2.01)	0.86	156	3.24 (2.33)	3.59 (2.15)	0.02*	0.27
Interpersonal strain	49	4.04 (3.45)	4.02 (3.28)	0.96	156	2.73 (2.52)	3.07 (2.94)	0.08	0.38
Demanding/manipulative	48	1.06 (1.39)	1.15 (1.52)	0.72	155	1.19 (1.61)	1.25 (1.51)	0.64	0.89

p¹=value for comparison between baseline and 1 year within case and control group

p²=value for comparison of change (from baseline to 1 year) between case and control

^*p<.05

IN-HOME-PD: Interdisciplinary home visits for Parkinson’s Disease; MCSI: Multidimensional Caregiver strain index; POP: Parkinson’s Outcomes Project

3.3. Feasibility analysis of Share the Care peer mentoring

We a recruited and trained 34 experienced caregivers as peer mentors. Peer mentors were primarily female (58.8%), White (79.4%), with a mean age of 63.6 years (SD 13.3) and a mean 8.4 years of caregiving (SD 6.4). Of the mentors enrolled and trained, 33 completed at least one and 19 completed two mentoring relationships. Fifty-one IN-HOME-PD caregivers began Share the Care at home visit 2 (86.4% enrollment rate) and 45 completed all 16 weeks of mentoring (88.2% retention rate). Reasons for withdrawal included PWPD death between visits 2-3 (n = 3), moving away (n = 1), and mentor withdrawal with mentee declining rematch (n = 3). Mentor-mentee pairs had a median of 12 calls (IQR 7) of 30 minutes’ duration (IQR 25) over 16 weeks. Mentees and mentors rated 100% of visits as useful and 100% reported getting along at each visit. Overall mean satisfaction was 90/100 for mentees and 91/100 for mentors on the CSI-SF after 16 weeks of mentoring. No mentees reported problems with their mentor or participation. Mentors reported issues with 14% of calls, including trouble reaching mentee by phone, difficulty understanding mentee’s accent, perception of mentee’s inability to sustain attention, and technological difficulties.

3.4. Secondary outcomes: Crossover effects of home visits with and without mentoring on caregiver strain, mental health, and self-efficacy

Using repeated measures ANOVA across all four IN-HOME-PD visits, we confirmed no significant change in caregiver strain at any study timepoint (p = 0.11), and the percentage of caregivers scoring in the mild, moderate, and severe strain categories did not shift over time (Table 4). Post-hoc pairwise comparisons showed that anxiety improved slightly between home visit 1 and 3 (7.07 (4.1) vs. 5.96 (3.95), p = 0.049). There were no changes in depression across the entire study period (p = 0.24) or within each inter-visit interval, nor in self-efficacy subscales (p = 0.90 and 0.62, respectively). Finally, peer mentors had no appreciable change in depression (2.1 (2.7) vs. 2.3 (2.6), p = 0.78), anxiety (4.0 (3.1) vs. 3.4 (2.6), p = 0.35), symptom management self-efficacy (39.6 (8.2) vs. 37.4 (10.2), p = 0.21), or community support services use (29.1 (8.3) vs. 29.1 (8.4), p = 0.97) between mentor training and the conclusion of their first 16-week mentoring relationship, though mentors as a group performed better in each category than mentees.

Table 4:

IN-HOME-PD Caregiver primary and secondary outcomes across home visits

Outcome measures (n = 51)	Visit 1	Visit 2	Visit 3	Visit 4	p value
Caregiver Strain (MCSI, range 0-72), mean (SD)	23.88 (9.41)	22.57 (9.14)	22.32 (9.63)	23.98 (9.72)	.11
Mild (0-19)	34.5%	41.4%	43.4%	36.7%
Moderate (20-29)	41.4%	36.2%	35.9%	44.9%
Severe (≥30)	24.1%	22.4%	20.8%	18.4%
Depression (HADS)	4.95 (3.46)	5.05 (3.75)	4.45 (3.42)	4.96 (3.54)	.24
Anxiety (HADS)	7.07 (4.1)	6.95 (3.9)	5.96 (3.95)	6.55 (3.84)	.049*
Self-Efficacy (CSES)
Symptom Management	35.88 (9.16)	36.76 (9.93)	36 (9.6)	36.14 (10.84)	.90
Community Support Services Use	24.25 (9.31)	25.54 (8.43)	25.59 (8.6)	26.1 (9.05)	.62

CSES: Caregiver Self-Efficacy Scale; HADS: Hospital Anxiety and Depression Scale; MCSI: Multidimensional Caregiver Strain Index. SD: Standard deviation.

4. Discussion

In this study testing the combined effects of one year of interdisciplinary telehealth-enhanced home visits plus 16 weeks of peer mentoring for caregivers of homebound PWPD, caregiver strain stabilized in the intervention group and worsened among a control group of caregivers despite the control PWPD having less severe disease. ^5,37 The acceptability and feasibility of both interventions were high among caregivers and peer mentors, and caregiver anxiety improved slightly, most notably during the period of peer mentoring. Thus, while IN-HOME-PD caregivers did not demonstrably improve, the interventions—individually or in combination—may have mitigated the course otherwise observed with usual care.

Despite their historic absence in the literature, recent data show that late-stage PD caregiver strain and depression worsen over time,⁴³ in line with studies demonstrating that the PWPD’s self-reported health⁴⁴ and motor severity⁴⁵ correlate with increasing caregiver strain, burden, anxiety, and depression. There are also notable sex differences in PD caregiving: in an analysis of 7,209 POP participants, caregivers of male PWPD had higher strain than caregivers of female PWPD (MCSI 19.9 vs. 16.4, p < 0.001). While POP caregiver demographics are not gathered, previous studies cite higher baseline and steeper increase in caregiver strain at ten years among female spousal caregivers, who comprised the majority of IN-HOME-PD caregivers and likely comprise the majority of POP caregivers of male PWPD.^19,46,47 As such, given that caregiver strain worsens longitudinally and perhaps especially among female caregivers and caregivers of late-stage and homebound PWPD, our findings of post-intervention stability are particularly noteworthy.

There are multiple potential explanations for caregiver strain mitigation in this multifaceted intervention. The Stress Process Model of Caregiving posits that strain results from caregiver and patient characteristics and stressors (e.g., PWPD’s cognition, mobility, dependency for activities of daily living, healthcare access, support system) and other stressors on the caregiver (e.g., financial and occupational demands, other caregiving responsibilities).^48-50 Strain leads to poorer caregiver physical and mental health, and worsening patient morbidity, institutionalization, and mortality.^{5,12,13,51-53} Knowledge, attitudes, caregiving mastery, and social support are critical mediators in the strain-caregiver outcomes and strain-patient outcomes pathways.^49,50,54 We hypothesize that the home visits and interim calls conferred additional knowledge and mastery to IN-HOME-PD caregivers than they would have received in usual care. Moreover, given the homebound status of their loved ones, usual care—as suggested by the near-absence of similarly-staged PWPD in POP—may have been limited or non-existent.²¹ Thus, education and continuity of care may have conferred knowledge and mastery, and peer mentoring offered structured, knowledgeable social support, that mitigated caregiver strain. Future studies will investigate the direct effect of peer mentoring on key mediators of caregiver strain—namely, knowledge, mastery, skills, and perceived social support—to determine whether these domains are effected in the short-term and if so, whether and over what time course these changes impact strain.

Peer mentoring interventions are grounded in the assumption that social support from a peer who has successfully navigated a similar health condition offers increased benefits over an individual’s existing social support or healthcare providers.⁵⁵ Peer mentoring has shown benefits in cancer, end-stage renal disease, and dementia caregiver populations, among others.^{23,26,28,56-59} Share the Care is the first peer mentoring intervention in PD and specifically focused on caregivers of homebound PWPD. Our findings mirror those of dementia caregiver peer mentoring studies.^57,60 While strain was stable, both PD caregivers and trained mentors had high adherence to and satisfaction with mentoring. We suspect that a driver of this successful implementation was the combination of practical knowledge, resources, and social support that mentors shared with their mentees. In a 2005 study by Brodaty among dementia caregivers with high strain, one third used no available community services, and another quarter used only one service.⁶¹ The majority of low-utilization caregivers felt that they did not need services, were unaware of or could not afford services, or their care recipient was reluctant. In future studies, we will explore the extent to which resource provision or recommendations directly from peer mentors overcome these barriers.

In the current study, mentoring was contingent on the PWPD-caregiver dyad qualifying for and consenting to both home visits and mentoring. This led to: 1) the inclusion of highly strained caregivers who, echoing Brodaty’s findings, felt mentoring was unnecessary and opted out, and 2) mentoring being limited to only caregivers of advanced, homebound PWPD in a geographically-limited area. Future standalone mentoring iterations will be more generalizable and incorporate community-based participatory research, leveraging caregivers’ input on key topics and resources in the intervention materials, and meaningful outcome measures.

There are several limitations of this study. Randomization of homebound dyads was neither ethically nor logistically possible: identifying and recruiting individuals definitionally at risk of loss to follow-up and randomizing them to usual care would have led to biased attrition, which is borne out in the lack of longitudinal follow-up of HY 4- & 5-staged POP PWPD nationally. Furthermore, both our prior home visit pilot⁶² and the current study included a significantly higher proportion of underrepresented minority PWPD than most published studies in the field, including those based on POP. As research recruitment and retention of underrepresented communities are paramount for improving representation and generalizability, and as stigma about research participation persists⁶³—randomization of doubly vulnerable homebound minority PWPD and their caregivers to observation without a guarantee about the provision or quality of usual care was deemed unethical. The second main limitation follows: the historical control group was notably non-equivalent, both in terms of baseline PWPD characteristics and caregiver strain, with the POP caregivers tending to healthier, younger, more mobile, and less cognitively impaired individuals. Yet, strain worsened longitudinally among the POP caregivers—as prior data predicts will occur with disease progression—but did not change among the intervention caregivers. One possible explanation is that homebound individuals in the intervention group were so advanced and their needs so great at the time of enrollment that minimal progression would be appreciated over one year, and thus caregiver strain might remain unchanged. However, as we previously reported,³⁴ IN-HOME-PD PWPD had significant worsening of the mobility and activities of daily living domains of the PDQ-39 (p = 0.01, Cohen’s d = 0.39; p = 0.03, d = 0.32, respectively, consistent with modest effect sizes), suggesting progression. This is despite a stabilization of overall quality of life compared to the significant worsening seen in the POP cohort (between-group difference in PDQ-39 total score, p = 0.04). Therefore, IN-HOME-PD PWPD and their caregivers began the intervention at a more severe baseline than controls in terms of stage, mobility, cognition, PWPD quality of life, and caregiver strain, experienced progression of stage and mobility over one year, and still demonstrated a stabilization of strain progression compared with controls. Peer mentoring was nested within and contingent on home visit participation, limiting generalizability and notably excluding caregiver experiences in earlier PD, those receiving care in the community, or those living in areas with fewer community resources and support systems. Additionally, the assumption of our power calculation—that MCSI would worsen by ten points—overestimated longitudinal change, leading to an underpowered study. The ideal duration of mentoring and timeline for measuring benefits may exceed the 16 weeks allotted here, and a more structured, resource-dense handbook grounded in behavior change theories may be more effective. In summary, further work is necessary to: identify whether our results are due to individual components of home visits, peer mentoring, or both; how each may be adapted to more sustainable, remote interventions; and how these interventions impact proximal domains in the causal pathway between PWPD and caregiver characteristics, strain, and dyadic outcomes. The retention and satisfaction of IN-HOME-PD dyads underscores the value of addressing these questions in future work.

5. Conclusions

Interdisciplinary telehealth-enhanced home visits combined with peer mentoring did not improve caregiver strain, but rather mitigated the decline observed in caregivers of less advanced individuals. Caregivers were amenable to and satisfied with both the labor-intensive home visits led by a team of healthcare professionals and the volunteer-led peer mentoring intervention. Effective interventions to help family caregivers are critically needed,¹⁷ particularly when paid caregiving is unattainable for most people.⁶⁴ Adapting key components of IN-HOME-PD visits and Share the Care peer mentoring into remote or asynchronous interventions may promote greater dissemination and sustainability. As the volume of family caregivers exponentially increases,¹ peer mentoring—already a successful model in other serious illnesses—may offer an adjunctive pipeline of support by leveraging the lived knowledge of experienced caregivers to benefit those who follow.

• Family caregivers of those with advanced Parkinson’s Disease experience high strain

• Home visits and peer mentoring in Parkinson’s caregivers mitigated worsening strain

• Caregiver peer mentoring had high retention, fidelity, utility, and satisfaction

Abbreviations

ANOVA

Analysis of variance

CSES

Caregiver Self-Efficacy Scale

CSI-SF

Client Satisfaction Inventory Short Form

HADS

Hospital Anxiety and Depression Scale

HY

Hoehn and Yahr Stage

IN-HOME-PD

Interdisciplinary Home Visits for Parkinson’s Disease

IQR

Interquartile range

MCSI

Multidimensional Caregiver Strain Index

PD

Parkinson disease

POP

Parkinson’s Outcomes Project

PWPD

Person with Parkinson’s Disease

SD

Standard deviation