Feasibility,acceptability,and pre- post-program outcome changes of the real-world implementation of an informal caregiver support program via text message

Abstract

Background

Few dementia caregiver support programs have been tested in real-world settings.

Objective

We tested the feasibility and pre- post-program outcome changes (e.g., preparedness for caregiving) of the first short message service texting program to support informal caregivers of people with dementia.

Methods

We analyzed observational data from 147 caregivers of people with dementia participating in a service program. This program was a remote, asynchronous, and bidirectional texting program focused on dementia education, skill-building, and community resources that lasted six months. We measured outcomes via surveys and metrics of program usage over six months.

Results

Two caregivers experienced technical issues during the program, 12.9% unsubscribed, and 68.8% read most texts thoroughly. Most caregivers (64.3–75.9%) reported the two highest levels of acceptability in four four-level items ranging from not at all to a lot/extremely. Levels of preparedness for caregiving and unmet needs improved from pre- to post-program testing in the full sample. We only observed pre- to post-program changes in depressive symptoms and strain when restricting the sample to those with worse baseline levels of those outcomes.

Conclusions

Acceptability and pre- to post-program changes in this real-world implementation of a texting caregiver support program were encouraging, though smaller than those observed in efficacy research trials, consistent with findings from other implemented caregiver programs.

Keywords

Alzheimer's disease caregiving dementia implementation mHealth

Introduction

Informal caregiving for people with dementia (PWD) is highly prevalent and leads to important negative health outcomes. Nearly 12 million people in the United States (US) provide unpaid care for a family member or friend with dementia.¹ Negative mental health outcomes of caregiving primarily involve distress, which is apparent in both self-reports and cortisol levels.² Caregivers of PWD are also more likely than non-caregivers and caregivers of other populations to have depression^3–6 and anxiety disorders.^4,5

Several efficacious caregiver support programs exist, including psychoeducational, cognitive behavioral therapy, counseling, support groups, PWD training, respite and multicomponent programs.⁷ However, many of the existing programs rely heavily on trained interventionists to deliver lengthy programs using rigid delivery modalities (mostly live and in-person) which challenge their implementation. Researchers have leveraged technologies to address implementation limitations. These technologies include websites focused on teaching caregiving coping skills,⁸ apps that deliver caregiver skills or mindfulness,⁹ and videocalls with multiple components (e.g., community resources, counseling, Alzheimer's disease and related dementias education).¹⁰ These are all promising programs, but relying only on these technologies may impact the reach of caregiving support programs, as these are not universally used.¹¹ In addition, most caregiver programs have been designed for urban, non-Latino Whites and results do not generalize to underserved groups.¹² Underserved groups have a higher risk of developing dementia, provide more intense and longer caregiving, experience higher levels of caregiver distress, depression, and burden, and are less likely to attend caregiver support services despite their interest in receiving support.^1,13,14 There is a need for easily implementable caregiver support solutions that do not leave anyone behind.

A potential solution to the barriers above-mentioned is the delivery of content tailored to different cultural and contextual groups via short message service (SMS) texting. SMS texting is highly accessible because nearly all adults nowadays have access to cell phones with SMS texting capabilities, and text messages are remote and asynchronous, meaning that time and space are not barriers to the delivery of caregiver support.^11,15 Addressing these accessibility barriers may increase reach, which is a crucial component of a program's population impact along with effectiveness.¹⁶ We had originally developed and shown the promising feasibility and preliminary efficacy outcomes of the first SMS texting program to deliver caregiver support to Latino informal caregivers, CuidaTEXT.^17,18 In 2022, we were approached to adapt CuidaTEXT to the broader community in a Midwest US region and implement it as a service. In this manuscript, we report the results from the implementation of this program (CareTEXT) in our region. We hypothesized that caregivers’ levels of strain would decline following their use of CareTEXT.

Methods

This is a secondary analysis of a pre-post-program observational assessment of one remote caregiver support quality improvement program that took place from August 2022 to June 2024 (no control group). One service/research institution administered CareTEXT to a convenience sample of caregivers in a five-county service region in the Midwest US. Recruitment sources included community events, social media, medical records and neurology clinic visits, research registries, word of mouth, and referrals from partner organizations and clinics. Caregivers’ eligibility criteria for the program included speaking English or Spanish, being at least 18 years old, the PWD living in one of the 5 eligible counties, owning a cell phone with a flat fee for texting, being able to read and write texts, emotionally or physically taking care of someone with a clinical or research diagnosis of dementia, and being willing to participate in the program. Due to the service-oriented and non-research priority of the program, we decided to stop asking the full survey (all demographic and pre-program outcomes) and only ask a basic satisfaction question to caregivers halfway through the program period. For this reason, the final analytic sample did not include all caregivers enrolled in the program (Figure 1). The program allowed more than one caregiver per PWD but did not identify those caregivers who cared for the same PWD. The University of Kansas Medical Center deemed this project to be a Quality Improvement project. All caregivers gave oral informed consent to participate in the program. This consent explained the program's goal and characteristics, its free and voluntary nature, the right not to respond to the surveys, and data collection and use procedures.

Program

CareTEXT is a six-month bilingual SMS remote and asynchronous program where caregivers receive scheduled messages and can also request on-demand assistance by texting. The original program (CuidaTEXT) was developed for Latinos. This program, and its development process, have been extensively detailed previously.¹⁷ CuidaTEXT is guided by the Stress Process Framework, which poses that caregivers’ background, context, and stressors have negative health consequences, and ameliorating factors of stress can reduce those stressors and its negative consequences.¹⁹ CuidaTEXT's delivery and content features were designed to engage stress ameliorating factors (e.g., preparedness for caregiving) while addressing common Latino context, background, and stress-related barriers, and in turn reduce the negative consequences of stress (e.g., depression).²⁰ Delivery features were integrated to avoid adding stressors and reduce context and background barriers. These included an asynchronous, largely automatic and remote modality with ability to choose preferred language and request on-demand messages, with language levels appropriate for people with low literacy. Content features were developed to address stressors while considering key context and background factors. These features included logistics, dementia education, self-care, social support, end-of-life care, managing dementia-related behaviors, problem-solving strategies, and community resources. Key context and background factors included cultural aspects, such as Spanish mock names for testimonials, videos with Latin dancing or recipes, or stress related to having family abroad. Message content also addressed key context and background factors by including links to community resources that are bilingual, remote, free, or that provide financial aid to address transportation, language, financial, and digital barriers.²⁰

Like CuidaTEXT, CareTEXT entails sending 1–3 automated daily messages covering the content features described above. Additionally, participants can text keyword-based queries for immediate assistance on the content features and engage in live chat sessions with the coach for further guidance upon request. The program had a single coach with a health-related bachelor's degree (who completed a master's degree during the period pf the program), had received brief training on dementia care from the team, and was guided by a protocolized manual to ensure consistency in tone, scope, and type of support provided across participants.

The research team adapted the program's message content to increase its acceptability among other frequent community groups in the region, namely [mostly urban] Black Americans and [mostly White] rural Americans. Work to conduct the adaptations followed Cultural Accommodation Model principles of co-creation with the target communities.²¹ For this reason, The research team held five, one-hour videocall meetings with five representatives of each of the two working groups to go over text messages and report whether they liked it the way it was or whether it required an adaptation. Adaptation suggestions were documented in a word file and were later discussed by the research team, who created a new version and presented it to the working groups to generate a final version. Working group participants made suggestions about adaptation to specific messages and the final adaptations were decided on site by consensus. Adaptations included reducing the number of names in Spanish and increasing names typical of other cultures, changing some expression and terms, adding new contexts (e.g., referencing the Deep South and the Church), and references and links to culturally-relevant lifestyle behaviors such as line dancing and soul foods.

Assessment

The research team collected information from three sources: baseline survey, six-month follow-up survey, and metrics of text message interactions. We collected sociodemographic and care-related information from the caregivers at baseline.

Outcomes included feasibility, acceptability, and pre- post-program outcome changes:

Feasibility outcomes were similar to those reported in our previous study.²² Outcomes included the feasibility of program enrollment (percentage of participants who were able to opt into CareTEXT), retention rate (completing follow-up survey), assessment (completing outcome scales), program delivery (technical issues), and program engagement (usage of texting features, unsubscribing, and reports of attention paid to CareTEXT).

Acceptability outcomes were also like the ones used in our previous study.²² We collected all of them in the six-month follow-up survey. These include four questions: The first three questions asked to what extent CareTEXT had “helped understand dementia more”, and “helped care for oneself” and “the PWD”. These questions had four response options ranging from not at all to a lot. A fourth question asked about overall satisfaction with CareTEXT, with four response options ranging from (1) not at all to (4) extremely.

Analyses of change in pre- post-program outcomes were exploratory, given the nature of the study. Pre- post-program outcomes included four scales about the caregiver. We decided to measure depressive symptoms and strain as negative consequences of caregiving in the Stress Process Framework, and caregiving competence as an ameliorating factor. We decided to also analyze an unmet needs checklist as a pragmatic service measure, as it allowed us to measure an important stressor. All scales have shown good psychometric properties previously in their original publications, except for the unmet needs checklist, which the research team developed:

Depressive symptoms (Center for Epidemiologic Studies Depression Scale [CES-D-10])²³: This is a 10-item, self-report rating scale that measures characteristic symptoms of depression in the past week (e.g., depression, loneliness). Each item is rated on a 4-point scale, from 0 (Rarely or None of the Time) to 3 (Most or All of the Time) with positively worded items (items 5 and 8) reverse scored. Items yield summary scores that range from 0 to 30, with higher scores indicating higher depression severity.

Caregiver strain (Modified Caregiver Strain Index [CSI])²⁴: The CSI is a 13-item screener that measures caregiver strain. For each of the items (e.g., caregiving is inconvenient), the caregiver can respond either No (0), Yes Sometimes (1), or Yes on a Regular Basis (2). The total score is the sum of all item scores. Higher scores indicate higher strain.

Caregiving competence (Preparedness for Caregiving Scale [PCS])^25,26: The PCS is a self-rated instrument that consists of eight items that ask caregivers how well prepared they believe they are for multiple domains of caregiving (e.g., caring for care recipient's physical needs). Responses are rated on a 5-point Likert scale with scores ranging from 0 (Not at All Prepared) to 4 (Very Well Prepared). The scale is scored by calculating the mean of all items answered, with a total score range of 0 to 4. The higher the score the more prepared the caregiver feels for caregiving.

Unmet needs checklist: This checklist was created by the contracting institution to inform the resources caregivers would need to get help with. This is a seven-item checklist for caregivers that requires 0 (No) or 1 (Yes) answers. The checklist states “In addition to the services you are already receiving, do you need any more services?”. The options include 1) connection to information and resources, 2) respite, 3) legal advice/assistance, 4) emotional/social support, 5) additional training, 6) home repairs or modifications to make caregiving tasks safer, and 7) other. A total score is obtained by summing all seven items. The higher the score, the higher the number of unmet needs the caregiver reports having unmet needs. The scale's concurrent validity using Spearman correlation with the CSI was 0.286; p < 0.01 (small to moderate effect size).²⁷

Analysis

We used descriptive statistics to summarize the baseline characteristics, and to report feasibility and acceptability outcomes. Given that we adapted the program to reflect specific racial, ethnic, and rural characteristics to enhance acceptability among underserved groups, we examined racial, ethnic, and rural differences in acceptability outcomes using chi-square tests. To assess pre- post-program outcome changes, we conducted paired-samples t-tests to evaluate changes from baseline to follow-up. For these t-tests, we also report effect sizes using Cohen's d's thresholds of small (0.2), medium (0.5), and large (0.8).²⁷ To reduce floor effects, we also examined pre- post-program outcome changes among participants with baseline scores corresponding to ≥25% and ≥50% of the possible score range for each scale exploratorily. We chose those baseline scores to use two comparable levels of severity across outcomes. We performed all analyses using SPSS Version 22.²⁸ The significance level was set at p < 0.05.

Results

Table 1 summarizes the baseline characteristics of caregivers. Out of the 301 individuals who enrolled in the program, we collected pre-program outcomes among 147 (final analytical sample) as this was a service-oriented program and the project needed to prioritize funds for enrolling people into the program and coaching versus conducting research. Those included were more likely to be White, rural, and to care for someone with a specific subtype of dementia versus unknown (see Supplemental Table 1). The final analytic sample (n = 147) averaged 63.6 years of age (SD 11.1; min 29; max 85) and 15.4 years of education (SD 2.5) and most reported being women (84.4%; n = 124), White (79.6%; n = 117), urban-living individuals (84.4%; n = 124), that were married or live with a partner (80.1%; n = 117), had no difficulties in paying for basic needs (12.4%; n = 18), were the partner of the PWD (51.0%; n = 75), and their PWD had either Alzheimer's disease (37.4%; n = 55) or an unknown dementia diagnosis (31.3%; n = 46). Completion of the follow-up feasibility and outcome scales assessment was more common among those who were older, were partners instead of adult children of the PWD, and in the case of outcomes alone, were more educated (Table 1).

Table 1.

Baseline characteristics of the caregivers being served in the CareTEXT program for the total sample, those who completed the follow-up survey outcomes (completers), and those who did not (non-completers).

Caregivers	Total (n = 147)	Feasibility completers (n = 113)	Feasibility non-completers (n = 34)	p value of feasibility completers vs. non-completers	Outcome scales completers (n = 100)	Outcome scales non-completers (n = 47)	p value of outcome scales completers vs. non-completers
Age in years, mean (SD)	63.6 (11.1)	64.9 (10.4)	59.4 (12.6)	0.01	65.2 (10.3)	60.1 (12.1)	0.01
Women, % (n)	84.4% (124)	82.3% (93)	91.2% (31)	0.21	83.0% (83)	87.2% (41)	0.51
Latino ethnicity, % (n)	8.2% (12)	8.8% (10)	5.9% (2)	0.73	9.0% (9)	6.4% (3)	0.75
Race, % (n)				0.38			0.34
White, % (n)	79.6% (117)	78.8% (89)	82.4% (28)		77.0% (77)	85.1% (40)
Black, % (n)	14.3% (21)	15.0% (17)	11.8% (4)		16.0% (16)	10.6% (5)
Other, % (n)	6.1% (9)	6.2% (7)	5.8% (2)		7.0% (7)	4.3% (2)
Rural residence, % (n)	15.6% (23)	15.0% (17)	17.6% (6)	0.71	14.0% (14)	19.1% (9)	0.42
Years of education, m (SD)	15.4 (2.5)	15.5 (2.3)	14.8 (3.0)	0.18	15.7 (2.3)	14.6 (2.7)	0.01
Without medical insurance, % (n)	2.7% (4)	2.7% (3)	3.0% (1)	0.49	2.0% (2)	4.3% (2)	0.53
Married or have a partner, % (n)	80.1% (117)	81.4% (92)	75.8% (25)	0.47	82.0% (82)	76.1% (35)	0.41
Difficult or very difficult to pay for basic needs, % (n)	12.4% (18)	9.7% (11)	21.9% (7)	0.08	9.0% (9)	20.0 (9)	0.06
Relation to PWD				0.01			0.01
Partner, % (n)	51.0% (75)	57.5% (65)	29.4% (10)		60.0% (60)	31.9% (15)
Children, % (n)	29.9% (44)	24.8% (28)	47.1% (16)		25.0% (25)	40.4% (19)
Other, % (n)	19.5% (28)	17.7% (20)	23.5% (8)		15.0% (15)	27.7% (13)
Dementia type of the PWD				0.23			0.15
Alzheimer's disease, % (n)	37.4% (55)	38.9% (44)	32.4% (11)		42.0% (42)	27.7% (13)
Unknown, % (n)	31.3% (46)	26.5% (30)	47.1% (16)		27.0% (27)	40.4% (19)
Mixed dementia, % (n)	9.5% (14)	12.4% (14)	2.9% (1)		10.0% (10)	10.6% (5)
Lewy bodies/Parkinson's dementia, % (n)	10.2% (15)	9.7% (11)	8.8% (3)		10.0% (10)	8.5% (4)
Other dementia, % (n)	11.6% (17)	12.4% (14)	8.8% (3)		11.0% (11)	12.8% (6)

Feasibility and acceptability findings can be seen in Table 2 and Figure 2, respectively. Out of the 147 caregivers who completed the full baseline survey, all were able to opt into CareTEXT, 113 (76.9%) completed at least the acceptability section of the follow-up survey and 100 (68.0%) completed the full survey, since their PWD was still alive. Ony two caregivers reported having [small] technical issues with CareTEXT, and these were related to messages not coming through at a certain point in time. Most caregivers (68.8%; n = 77) read most messages thoroughly, 12.9% (n = 19) unsubscribed from the program before it ended, and 81.0% (n = 119) sent at least one text message to the program. The reasons for unsubscribing included their loved one passing away or entering long-term care (n = 5), receiving too many messages (n = 5), and messages not applying to them (n = 1). Most participants perceived CareTEXT helped them very or extremely understand dementia more (75.9%); and care for themselves (66.3%), care for their PWD (70.6%), and most were very or extremely satisfied with CareTEXT (73.2%). These outcomes did not vary by rurality, and only one of these differed by race. Even though 68.2% of White caregivers reported being very or extremely satisfied with the program, this group were the only ones who reported no satisfaction at all (5.7% within that racial group), and 97.0% of those who reported being somewhat satisfied with the program.

Table 2.

Feasibility of CareTEXT.

Feasibility
Program enrollment	Number of caregivers enrolled into CareTEXT	n = 301
	Percentage of pre-program outcome respondents able to opt into CareTEXT^a	100% (n = 147)
Retention rate	Percentage of pre-program outcome respondents who completed the follow-up survey^a	76.9% (n = 113)
Assessment feasibility	Percentage of pre-program outcome respondents who completed the outcome scales^a	68.0% (n = 100)
Program delivery	Percentage of feasibility outcome completers who experienced technical issues: First message took a few days to come through, 2) some issues with reception of messages^b	1.4% (n = 2)
Program engagement	Percentage of pre-program outcome respondents who unsubscribed from CareTEXT^a	12.9% (19)
	Percentage of pre-program outcome respondents who sent 0 text messages (no interaction)^a	19.0% (n = 28)
	Percentage of pre-program outcome respondents who sent 1–9 text messages (low interaction)^a	38.1% (n = 56)
	Percentage of pre-program outcome respondents who sent 10–49 text messages (intermediate interaction)^a	32.7% (n = 48)
	Percentage of pre-program outcome respondents who sent 50–99 text messages (high interaction)^a	7.5% (n = 11)
	Percentage of pre-program outcome respondents who sent more than 100 text messages (very high interaction)^a	2.7% (n = 4)
	Percentage of feasibility outcome completers who read through most messages thoroughly^b	68.8% (77)
	Percentage of feasibility outcome completers who took a short look at messages most times^b	30.4% (34)
	Percentage of feasibility outcome completers who did not read messages most times^b	0.9% (1)

^acalculated for pre-program outcome respondents (n = 147); ^bcalculated for feasibility outcome completers (n = 113).

Pre- post-program outcome changes can be seen in Table 3. For the full sample, participation in CareTEXT was not followed by an improvement in caregiver depression as rated with the CES-D-10 or strain as rated with the CSI. However, two outcomes improved after participating in CareTEXT. Outcomes that improved included preparedness for caregiving as measured by the PCS, which increased from 2.3 to 2.5 (95% confidence interval of mean difference [95% CI] = −0.34; −0.09; Cohen's d = 0.33), and unmet needs, which decreased from 1.7 to 0.9 (95% CI of mean difference = 0.51; 1.05; Cohen's d = 0.56). Depressive symptoms improved among those scoring at least a 7 out of 30 at baseline, from 11.7 to 9.9 (95% CI of mean difference = 0.63; 2.92; Cohen's d = 0.38). Strain improved only among those scoring at least 12 out of 24, from 16.3 to 13.9 (95% CI of mean difference = 1.06; 3.60; Cohen's d = 0.62). Effect sizes increased in all outcomes as we gradually restricted the sample to those with worse baseline scores in that outcome.

Table 3.

Pre- post-program outcome changes.

Caregiver	n	M baseline	M follow up	M difference	SD of M difference	95% CI lower	95% CI upper	P value	Cohen's d
CES-D-10 depression (0–30)	95	9.3	8.6	0.69	5.00	−0.32	1.71	0.12	0.14
If baseline CES-D-10 score ≥ 7	67	11.7	9.9	1.78	4.69	0.63	2.92	<0.01	0.38
If baseline CES-D-10 score ≥ 15	15	17.8	13.2	4.60	4.26	2.24	6.96	<0.01	1.08
CSI Strain (0–24)	97	10.5	11.3	−0.77	4.53	−1.68	0.15	0.10	0.17
If baseline CSI score ≥ 6	82	11.8	11.9	−0.09	5.30	−1.03	0.90	0.86	0.02
If baseline CSI score ≥ 12	36	16.3	13.9	2.33	3.75	1.06	3.60	<0.01	0.62
PCS Preparedness (0–4)	100	2.3	2.5	−0.21	0.65	−0.34	−0.09	<0.01	0.33
If baseline PCS score ≤ 3	89	2.1	2.4	−0.30	0.55	−0.42	−0.18	<0.01	0.55
If baseline PCS score ≤ 2	39	1.6	2.1	−0.53	0.52	−0.70	−0.36	<0.01	1.02
Unmet needs (0–7)	100	1.7	0.9	0.8	1.38	0.51	1.05	<0.01	0.56
If baseline unmet needs score ≥ 2	35	3.1	1.1	2.00	1.43	1.51	2.49	<0.01	1.13

Discussion

In this study, we described the feasibility, acceptability and pre- post-program outcome change findings from a real-world implementation of CareTEXT among informal caregivers. This is the first time an SMS texting program for dementia caregiving support has been implemented in the real world. Results suggest that CareTEXT can be implemented feasibly and with most caregivers reporting high levels of acceptability in the real world among people of different ethnic, racial and urban/rural setting backgrounds. Rather than decreasing depressive symptoms or strain, CareTEXT might be more likely to increase caregiver preparedness for caregiving and reduce unmet needs in this population.

Our findings add to the scarce literature of dementia caregiver support real-world program implementation. Only 15 dementia caregiver support programs have been tested in the real world.^29,30 Some of these include well-established programs such as the Resources for Enhancing Alzheimer's Caregiver Health II, Reducing Disability in Alzheimer's Disease, and Savvy Caregiver.^31–33 To our knowledge, this is the only dementia caregiver support program that relies largely on SMS texting. SMS texting is readily available to most caregivers given that most adults own a cell phone with SMS texting capabilities, and SMS texting's asynchronous nature allows caregivers to access the content of the program at anytime and anywhere.^11,15

Some feasibility and acceptability findings contrast with previous versions of the current SMS texting program and SMS texting programs for other conditions. CareTEXT caregivers had a lower engagement with text messages than previous versions. For example, previously, 25% of caregivers texted 100 messages or more during the six months of the program. However, in the current study, the percentage was 2.7%. Similarly, 86% reported reading most texts thoroughly in the previous study versus 69% of our current sample. A cause for the higher engagement in the previous study might be the inclusion of a largely non-Latino and older sample. Latinos and younger generations engage more with SMS texting than other populations,¹⁵ and Latinos’ baseline dementia knowledge is lower.³⁴ While unsubscription rates were higher than our previous study (13% versus 0%), they were still lower than texting programs for other conditions that were not co-developed with the target audience (e.g., 34% within 4 weeks).³⁵ It is important to consider that this is a real-word implementation, and therefore, the sample is not as biased towards those highly motivated in participating as research samples. In fact, real-world implementation of caregiver support programs that relied on in-person meetings has also achieved lower program-engagement. For example, a local adaptation of Reducing Disability in Alzheimer's Disease achieved 63% completion of nine out of the 12 modules of the program.³¹

Participation in CareTEXT was accompanied by observations of improved pre- post-program outcome changes. These outcomes included preparedness for caregiving and unmet needs in the total sample, and depressive symptoms and strain, only when restricting the sample to those with worse baseline levels of that outcome. This higher sensitivity of preparedness for caregiving and unmet needs might be due to these being more proximal factors (stressors and ameliorating factors) than depressive symptoms and strain (negative consequences of stressors) in the Stress Process Framework.¹⁹ Effect sizes for the full sample were null for depressive symptoms and strain, small for preparedness for caregiving and medium for unmet needs. However, the effect size of all outcome changes increased gradually as the samples were restricted to those with worse baseline scores, reaching moderate and high effect sizes in the worst-off samples. Preparedness for caregiving increased after using CareTEXT, which is in line with its original study with the Latino population.¹⁸ Improving preparedness for caregiving is important because it has been described as a target for dementia caregiving support programs, given its association with improvement with multiple other caregiver and care recipient mental health and safety outcomes.³⁶ Our study resulted in the same percentage of observed reduction of unmet needs as a real-world dyad home-based behavior management and exercise program (47%).³¹ Another real-world implementation of the same home-based program observed similar results, including a 35% reduction in unmet needs.³⁷ The lack of a change on caregiver depressive symptoms and strain in the full sample is not in line with our previous version of CareTEXT in the Latino population but aligns with other caregiver programs. For example, telephone-based support groups and a video-phone psychosocial program led to no improvements in depressive symptoms six months after baseline.^10,38 These similar results might be explained by the fact that none of these two studies restricted the sample to those with some baseline level of depressive symptoms. In fact, another technology-based caregiver support trial only found statistically significant differences with the control group when restricting the sample to those with high depression at baseline.³⁹ CareTEXT was developed for Latinos and adapted for [mostly urban] Black Americans and [mostly White] rural Americans. It is possible that outcome changes could have been more pronounced had the sample reflected these communities, since ∼80% of our sample identified as White, 91.8% as non-Latino, and 84.4% as urban.

The current study has limitations. This was a quality improvement research project of a service program, which might reduce the generalizability of findings due to its local nature and non-research priority focus. Disrupting the assessment of feasibility, acceptability and pre- post outcomes early (after the first 147 participants) for funding-related reasons might have biased the outcomes, as those for whom this information was not continued be collected were more likely to be from racialized and urban communities. For example, the slightly lower satisfaction with CareTEXT among White compared to non-White caregivers might have resulted in a higher overall satisfaction had the assessment disruption not taken place. We did not include a control group and therefore cannot attribute causality of improvements in outcomes to participating in the program. The program allowed more than one caregiver per PWD but did not identify those caregivers who cared for the same PWD. Doing so may have introduced non-independence of observations and shared contextual influences that could affect the variability and precision estimates. However, in our past research, we found that having more than one caregiver per PWD participating in CuidaTEXT is rare, despite being encouraged.¹⁸ We did not report fidelity, which is important for the implementation of programs in the real-world.³⁰ However, team leaders met weekly and as needed with the coach to check in on message interactions and troubleshoot any issues. Since the implementation relied on the same team as the research, we also did not analyze other implementation science outcomes such as the adoption or reach of interventionists. However, our program implementation ended along with the end of funding, which indicates a reliance on funding for sustainability, mostly to fund the time for enrollment and coach support. It is important to highlight that, even if the program ended, caregivers still have access to all the messages received (including the contact information of 24/7 free resources), and the last messages they received from the program invited them to reflect on what they learned and contact these resources when needed. There was a 23% non-completion rate of follow-up surveys. These individuals could have had more undesirable feedback that was not assessed than those who completed the follow-up surveys. The unmet needs checklist lacks strong validation and should be interpreted cautiously.

This study has implications for public health and future research. Researchers might want to do more formative work to increase the levels of acceptability, especially to increase levels of satisfaction among individuals identifying as White. To continue implementing this program, it is crucial either to identify funding mechanisms, or to develop fully automatic versions of this program to offset the cost of coaches. A solution for dementia-focused research centers could be to fund this program as an engagement tool, by which, besides supporting families, they also promote dementia research studies. A previous fully powered efficacy study is more pressing than implementation, as this has not yet been established. Efficacy studies will need to restrict the eligibility criteria to only those with some level of baseline depressive symptoms and/or strain to potentially detect differences, if these are primary outcomes. Once efficacy has been established, studies could also assess the population impact of this versus in-person programs to consider effectiveness and reach. Future studies should analyze the fidelity of the program. While this analysis is outside the scope of the current study, we have tracked all interactions between the caregivers and CareTEXT automatic and coach-based messages. Future studies should also assess barriers and facilitators of implementing CareTEXT in other settings such as dementia associations, aging agencies and health care settings.

Conclusion

This work contributes to the scarce literature on real-world implementation of caregiver support programs, by reporting results from the first SMS texting caregiver support program. Findings indicate that, as soon as CareTEXT shows efficacy, it can be implemented with a certain level of feasibility, acceptability and changes in pre- post-program outcomes. Future research is warranted to increase levels of acceptability, especially among those identifying as White, to assess its efficacy, and to better understand barriers and facilitators to implement this program in other settings and to maximize sustainability.

Supplemental Material

sj-docx-1-alz-10.1177_13872877261450936 - Supplemental material for Feasibility, acceptability, and pre- post-program outcome changes of the real-world implementation of an informal caregiver support program via text message

Supplemental material, sj-docx-1-alz-10.1177_13872877261450936 for Feasibility, acceptability, and pre- post-program outcome changes of the real-world implementation of an informal caregiver support program via text message by Jaime Perales-Puchalt, Christina Baker, Binod Wagle, Melissa Godar, Sayra Nieto Gomez, Hayden Johnson, Betty Drees, Mónica Fracachán-Cabrera and Mariana Ramírez-Mantilla in Journal of Alzheimer's Disease

Footnotes

Acknowledgements

JPP thanks the national and local organizations that have partnered with him to conduct present and past research since 2015. Especially the Mid-America Regional Council, James Stowe and Sarah Albin for letting us know about the funds and guiding us through the bureaucracy. The research team thanks the Community Advisory Board who helped adapt the program to diverse populations and all families included in all stages of this project as well as anyone who has contributed directly and indirectly to it. We also thank all platforms and organizations which have shared the opportunity to participate in this project with the people they serve. The authors used ChatGPT (OpenAI) to assist with language editing and clarity. The authors reviewed and take full responsibility for the final content. The ideas and opinions expressed herein are those of the authors alone, and endorsement by the authors’ institutions or the funding agency is not intended and should not be inferred.

ORCID iD

Jaime Perales-Puchalt

Ethical considerations

The University of Kansas Medical Center deemed this project as conducted as a Quality Improvement initiative, and as such was not formally supervised by an Institutional Review Board.

Consent to participate

All caregivers gave oral informed consent to participate in the program. This consent explained the program's goal and characteristics, its free and voluntary nature, the right not to respond to the surveys, and data collection and use procedures.

Consent for publication

Author contribution(s)

Jaime Perales-Puchalt: Conceptualization; Data curation; Formal analysis; Funding acquisition; Investigation; Methodology; Project administration; Supervision; Writing – original draft; Writing – review & editing.

Christina Baker: Investigation; Project administration; Resources; Writing – review & editing.

Binod Wagle: Data curation; Project administration; Resources; Supervision; Writing – review & editing.

Melissa Godar: Data curation; Investigation; Project administration; Resources; Writing – review & editing.

Sayra Nieto Gomez: Data curation; Investigation; Project administration; Resources; Writing – review & editing.

Hayden Johnson: Data curation; Project administration; Resources; Writing – review & editing.

Betty Drees: Data curation; Methodology; Resources; Supervision; Writing – review & editing.

Mónica Fracachán-Cabrera: Data curation; Project administration; Resources; Supervision; Writing – review & editing.

Mariana Ramírez-Mantilla: Conceptualization; Investigation; Methodology; Project administration; Resources; Supervision; Writing – review & editing.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Mid-America Regional Council. JPP's time was also supported by Grants K01 MD014177 and P30 AG072973 from the NIH.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data availability statement

Data are not publicly available because this project was conducted as a quality improvement initiative rather than a research study, and data sharing was not included in the original scope or consent procedures

Supplemental material

Supplemental material for this article is available online.

References

Alzheimer's Association. 2024 Alzheimer's disease facts and figures. Alzheimers Dement 2024; 20: 1–146.

Corrêa

Vedovelli

Giacobbo

, et al. Psychophysiological correlates of cognitive deficits in family caregivers of patients with Alzheimer disease. Neuroscience 2015; 286: 371–382.

Cuijpers

. Depressive disorders in caregivers of dementia patients: a systematic review. Aging Ment Health 2005; 9: 325–330.

Sallim

Sayampanathan

Cuttilan

, et al. Prevalence of mental health disorders among caregivers of patients with Alzheimer disease. J Am Med Direct Assoc 2015; 16: 1034–1041.

Atteih

Mellon

Hall

, et al. Implications of stroke for caregiver outcomes: findings from the ASPIRE-S study. Int J Stroke 2015; 10: 918–923.

Thunyadee

Sitthimongkol

Sangon

, et al. Predictors of depressive symptoms and physical health in caregivers of individuals with schizophrenia. Nurs Health Sci 2015; 17: 412–419.

Walter

Pinquart

. How effective are dementia caregiver interventions? An updated comprehensive meta-analysis. Gerontologist 2020; 60: e609–e619.

Waller

Dilworth

Mansfield

, et al. Computer and telephone delivered interventions to support caregivers of people with dementia: a systematic review of research output and quality. BMC Geriatr 2017; 17: 265.

Jain

Okereke

Gitlin

, et al. Mentalizing imagery therapy to augment skills training for dementia caregivers: protocol for a randomized, controlled trial of a mobile application and digital phenotyping. Contemp Clin Trials 2022; 116: 106737.

10.

Czaja

Loewenstein

Schulz

, et al. A videophone psychosocial intervention for dementia caregivers. Am J Geriatr Psychiatry 2013; 21: 1071–1081.

11.

Pew Research Center. Mobile fact sheet, https://www.pewresearch.org/internet/fact-sheet/mobile/ (2021, accessed 2021).

12.

Gilmore-Bykovskyi

Johnson

Walljasper

, et al. Underreporting of gender and race/ethnicity differences in NIH-funded dementia caregiver support interventions. Am J Alzheimers Dis Other Demen 2018; 33: 145–152.

13.

Weden

Shih

Kabeto

, et al. Secular trends in dementia and cognitive impairment of US rural and urban older adults. Am J Prev Med 2018; 54: 164–172.

14.

Hudomiet

Hurd

Rohwedder

. Trends in inequalities in the prevalence of dementia in the United States. Proc Natl Acad Sci U S A 2022; 119: e2212205119.

15.

Duggan

. Cell phone activities 2013. Washington, DC: Pew Research Center, 2013. https://www.pewresearch.org/internet/2013/09/19/cell-phone-activities-2013/ (2013).

16.

Rose

. Sick individuals and sick populations. Int J Epidemiol 2001; 30: 427–432.

17.

Perales-Puchalt

Acosta-Rullan

Ramirez-Mantilla

, et al. Development of CuidaTEXT: a text message intervention to support Latino dementia family caregivers. medRxiv 2021: 2021.2012.2006.21267369.

18.

Perales-Puchalt

Ramírez-Mantilla

Fracachán-Cabrera

, et al. A text message intervention to support latino dementia family caregivers (CuidaTEXT): feasibility study. Clin Gerontol 2024; 47: 50–65.

19.

Pearlin

Mullan

Semple

, et al. Caregiving and the stress process: an overview of concepts and their measures. Gerontologist 1990; 30: 583–594.

20.

Perales-Puchalt

Ramírez-Mantilla

Moore

, et al. Testing the effects of a text message intervention on depression and distress among Latino dementia caregivers: a randomized controlled trial protocol. Contemp Clin Trials 2026; 161: 108214.

21.

Burrow-Sanchez

Martinez

Hops

, et al. Cultural accommodation of substance abuse treatment for Latino adolescents. J Ethn Subst Abuse 2011; 10: 202–225.

22.

Perales-Puchalt

Checa

Espejo

, et al. Validation of wellbeing scales among informal caregivers of Latinos with Alzheimer’s disease and related dementias. Hisp J Behav Sci 2025; 47: 10.1177/07399863241308996.

23.

Cheng

Chan

. The center for epidemiologic studies depression scale in older Chinese: thresholds for long and short forms. Int J Geriatr Psychiatry 2005; 20: 465–470.

24.

Thornton

Travis

. Analysis of the reliability of the modified caregiver strain index. J Gerontol B Psychol Sci Soc Sci 2003; 58: S127–S132.

25.

Carter

Stewart

Archbold

, et al. Living with a person who has Parkinson's disease: the spouse's perspective by stage of disease. Mov Disord 1998; 13: 20–28.

26.

Gutierrez-Baena

Romero-Grimaldi

. Development and psychometric testing of the Spanish version of the caregiver preparedness scale. Nurs Open 2021; 8: 1183–1193.

27.

Cohen

. Statistical power analysis for the behavioral sciences. New York: Routledge, 2013.

28.

IBM Corp. Released 2023. IBM SPSS Statistics for Windows, Version 29.0. Armonk, NY: IBM Corp., 2023.

29.

Hodgson

Gitlin

. Chapter 6 - implementing and sustaining family care programs in real-world settings: barriers and facilitators. In: Gaugler

(eds) Bridging the family care gap. London: Academic Press, 2021, pp.179–219.

30.

Hwang

Hodgson

Gitlin

. Implementing dementia caregiver programs in real-world settings: fidelity considerations. J Am Med Dir Assoc 2024; 25: 34–40.e11.

31.

Perales-Puchalt

Barton

Ptomey

, et al. Effectiveness of “reducing disability in Alzheimer’s disease” among dyads with moderate dementia. J Appl Gerontol 2021; 40: 1163–1171.

32.

Burgio

Collins

Schmid

, et al. Translating the REACH caregiver intervention for use by area agency on aging personnel: the REACH OUT program. Gerontologist 2009; 49: 103–116.

33.

Griffiths

Whitney

Kovaleva

, et al. Development and implementation of tele-savvy for dementia caregivers: a department of veterans affairs clinical demonstration project. Gerontologist 2015; 56: 145–154.

34.

Connell

Roberts

McLaughlin

. Public opinion about Alzheimer disease among blacks, hispanics, and whites: results from a national survey. Alzheimer Dis Assoc Disord 2007; 21: 232–240.

35.

Abroms

Hershcovitz

Boal

, et al. Feasibility and acceptability of a text messaging program for smoking cessation in Israel. J Health Commun 2015; 20: 903–909.

36.

Hancock

Czaja

Schulz

. The role of preparedness for caregiving on the relationship between caregiver distress and potentially harmful behaviors. Am J Alzheimers Dis Other Demen 2022; 37: 15333175221141552.

37.

Menne

Bass

Johnson

, et al. Statewide implementation of “reducing disability in Alzheimer’s disease”: impact on family caregiver outcomes. J Gerontol Soc Work 2014; 57: 626–639.

38.

Martindale-Adams

Nichols

Burns

, et al. A trial of dementia caregiver telephone support. Can J Nurs Res Archive 2013: 30–49.

39.

Finkel

Czaja

Martinovich

, et al. E-care: a telecommunications technology intervention for family caregivers of dementia patients. Am J Geriatr Psychiatry 2007; 15: 443–448.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.03 MB