Telehealth Interventions to Support Self-Management in Stroke Survivors: A Systematic Review

Telehealth (TH) intervention is a method to optimize self-management (SM) support in stroke survivors. Objectives of this study included identifying the TH-SM intervention’s focus and SM support components, the TH delivery type, and the TH-SM support effects on stroke survivors. Five databases were searched for the years 2005–2020 to identify TH-SM support interventions for stroke survivors. Randomized controlled trials and quasi-experimental, one-group re-post study designs were included. Ten studies were reviewed. TH-SM support focused on post-stroke depression, obesity management, participation, functional mobility, and activities of daily living. The TH delivery type most used in selected studies was messaging. Regarding the SM support components, the education component was used in all studies, and psychological support and lifestyle advice and support were used in 8 out of 10 studies. TH-SM intervention had positive effects in terms of goal achievement for SM behavior, emotional state, and mobility of clinical outcomes, and TH acceptance in stroke survivors. Although the TH-SM-supported intervention effects were not found consistently in all outcomes, this review discovered a positive effect on various SM-related outcomes. In addition, TH delivery types and SM support components showed the possibility of various options to be considered for intervention. Therefore, we suggest that TH-SM supported intervention is a positive alternative for SM support in stroke survivors.


Introduction
In stroke care, emphasis has been placed on intense medical care and rehabilitation benefits in the acute phase, but care and support in the subsequent stages are as important as acute phase care [1]. People experiencing long-term health conditions, such as stroke survivors, must learn new behaviors or adjust their lifestyle to suit their needs [2].
Improving self-management (SM) is an important challenge for healthcare systems worldwide for long-term condition (LTC) survivorship [3]. According to the U.S. Institute of Medicine, SM is defined as tasks that one must perform while living with one or more chronic conditions [4]. The main elements are medical and behavioral management, emotional management, and role management, and the SM intervention empowers clients by providing them with the knowledge and skills to manage these tasks well [5]. For effective SM support implementation, Pearce et al. [6] developed the Practical Reviews in Self-Management Support (PRISMS) taxonomy, which consists of 14 components for SM support intervention in LTCs. Although many interventions use multiple components of SM support, the effective configuration and implementation of SM components is crucial [7]. Based on the PRISMS taxonomy, Hanlon et al. [8] proposed six components for SM support using telehealth (TH): patient education and information, monitoring using Healthcare 2021, 9, 472 2 of 19 feedback and action plans, clinical reviews, adherence support, psychological support, and lifestyle interventions.
TH provides an appropriate alternative in healthcare, where the provision of acceptable and high-quality primary care is limited due to the increase in life expectancy and consequent chronic conditions [9]. It helps clients manage their condition through improved self-care and access to education and support systems. In addition, clients and health professionals can remotely exchange important clinical information for the management and support of LTCs [10]. Telerehabilitation has been conducted using telecommunication devices to provide evaluations and interventions for improving the motor, cognitive, and psychosocial function of stroke survivors [11,12]. It has become a means to meet the rehabilitation needs of stroke survivors living in rural areas and low-and middle-income countries, where stroke care is burdensome [13,14]. The increasing use of mobile phones and personal computers/tablets can be a valuable resource for lifestyle change and disease management. Specifically, the online eHealth support tool, accessible from a handheld device or personal computer, is one way to optimize SM support for clients' goals [15,16].
With the background of various TH interventions and techniques, we aimed to obtain a broad overview of the evidence of TH-SM support in stroke survivors using a systematic review methodology. Therefore, the objectives of this review were to: (1) identify the focus of TH-SM intervention and SM support components in stroke survivors, (2) identify the type of TH delivery, and (3) identify the effects of the TH-SM support in stroke survivors.

Search Strategy and Selection Criteria
We searched five databases for reviews (Medline Complete, Embase, CINAHL, PsycINFO, and Web of Science) to identify relevant studies published on the TH-SM support interventions for stroke survivors between 2005 and 2020. Our basic search strategy included "telehealth terms" and "stroke terms" and "self-management support terms." The following keywords were used: "telehealth," "e-health," "mobile health," "telerehabilitation," "telecare," "telehealthcare," and "stroke," "cerebrovascular accident," "CVA," "cerebrovascular stroke," "cerebrovascular apoplexy," and "self-care," "self-management," "self-help," "self-monitor," "lifestyle," "patient education." The eligibility criteria included: (1) individuals with a clinical diagnosis of stroke, (2) interventions focused on SM support, (3) studies using TH intervention, including telephone calls and/or web-or mobile-app-based interventions, and (4) studies with outcomes focused on goal attainment, self-care, activities of daily living, self-efficacy, quality of life, lifestyle behavior, participation, function, psychosocial, disability, and adherence. We excluded: (1) non-English publications, (2) single SM strategy interventions (a study focused on a specific strategy as opposed to interventions using broader strategies), (3) Studies involving only caregiver-related outcomes, and (4) dissertations, theses, and protocol studies. The study titles and abstracts were examined after initial search. Furthermore, we obtained the full text of eligible studies based on the eligibility and exclusion criteria. The manuscript was searched for eligibility and exclusion criteria. A consensus was sought between the authors for the study to be included in the systematic review. This review was guided by the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement [17].

Quality Appraisal
Among the selected studies, the quality of randomized controlled trials (RCT) were appraised using the Physiotherapy Evidence Database (PEDro) Scale [18]. Three of the four studies showed "high" quality. The items with low scores on the scale used blind subjects and therapists ( Table 1). The Risk of Bias Assessment tool for Non-randomized Studies (RoBANS) was used to appraise the quality of six non-randomized controlled trials (NRCT) [19]. In the comparability of participants, one study was evaluated as having a high risk of bias due to the pre-and post-design of the non-equivalent control group whose Healthcare 2021, 9, 472 3 of 19 homogeneity between the intervention group and the control group was not verified. Four studies showed a high risk of bias in the confounding variables, as there was no group to control for confounding variables or a control group whose homogeneity was not verified. As none of the six studies mentioned evaluator blinding, the risk of bias was evaluated in terms of blinding the outcome assessment. Items other than these indicated a low risk of bias (Table 2).

Outcomes and Relevance
In this review, the following primary outcomes of the TH-SM intervention were targeted: (1) SM behaviors, goal achievement, activities of daily living (ADL), participation, and medication adherence; (2) clinical outcome; disability or recovery level, physical function, mobility, fatigue, and emotion; (3) self-efficacy; (4) quality of life (QOL); and (5) TH acceptance, adherence, subjective feedback, and satisfaction.

Results
Of the 1028 titles and abstracts, ten studies were included in this review of TH-SM interventions for stroke survivors [20][21][22][23][24][25][26][27][28]. Figure 1 shows a PRISMA flow chart shown in Figure 1 illustrates the search results and review selection. Two reviewers screened the full texts independently. The data were completed by one reviewer using a data extraction  Of the 1028 titles and abstracts, ten studies were included in this review of TH-SM interventions for stroke survivors [20][21][22][23][24][25][26][27][28]. Figure 1 shows a PRISMA flow chart shown in Figure 1 illustrates the search results and review selection. Two reviewers screened the full texts independently. The data were completed by one reviewer using a data extraction table, subsequently checked by a second reviewer for accuracy, with discrepancies resolved by discussion.

Participants' Characteristics
A total of 427 participants were included in ten studies. Sample sizes varied throughout the studies, ranging from 12 [26] to 147 participants [24]. The phases of recovery after participants' stroke was acute phase (n = 2), chronic phase (n = 1), and both acute and chronic survivors for 3-8 months or less than 24 months (n = 5). One study did not mention the time course (Table 3).

Participants' Characteristics
A total of 427 participants were included in ten studies. Sample sizes varied throughout the studies, ranging from 12 [26] to 147 participants [24]. The phases of recovery after participants' stroke was acute phase (n = 2), chronic phase (n = 1), and both acute and chronic survivors for 3-8 months or less than 24 months (n = 5). One study did not mention the time course (Table 3).

Types of TH delivery
Messaging was the most common in the selected studies, provided through short message service (SMS), email, in-home messaging devices (IHMDs), and smartphone push notification. All but one study [20] using messaging services combined with phone calls or web-based [29] or app-based interventions [23]. Four studies used phone calls [21,22,27,28], while four others used web-based intervention, such as video conferencing [24][25][26] or a web platform [24,29]. One study used an app platform for smart devices [23]. The TH types were synchronous (n = 2), asynchronous (n = 2), or both (n = 6) ( Table 3).

Frequency, Duration, and Length of Intervention
The frequency of intervention varied according to the characteristics of the intervention sessions. Messaging was used at a frequency of once a day [20][21][22]29] or twice a day [28], or once a week [23]. Telephone sessions were once a week [27], twice a week [28], or five times throughout the course of the intervention [21,22]. The video conferencing session was performed twice a week [25,26], or four times during the intervention period [24]. Three studies reported the duration of each phone call or video conferencing session as 5-15 min [27], 20 min [24], and 2 h [25]. In some studies, the exact duration of the intervention was not mentioned [20][21][22]28,29]. The length of interventions ranged from 1 to 6 months (Table 3).

Intervention
The focus of the TH-SM interventions varied. Cadilhac et al. [20] and Kamoen et al. [24] focused on lifestyle behavior changes and disease management. The remaining studies explored the relief of depressive symptoms after stroke [27], physical activity and participation [25,26], functional mobility [21,22], obesity management [23], and ADL [28,29] ( Table 3). Most studies excluded stroke survivors with severe aphasia and cognitive decline, but one study included participants with cognitive decline if they could participate in the intervention with the help of their caregiver. The majority of interventions were home-based, while other studies included delivery in a specific place in the area where the participants lived, prepared in advance by researchers.
Various SM support components were used in selected studies, and we analyzed the regrouped components from the PRISMS taxonomy of SM support for LTCs. Table 4 lists the SM support components analyzed in this study. All studies used at least four of nine components.  Education components were used in all ten studies. They addressed personal recovery and prevention post-stroke; level of functional ability; post-stroke depression (PSD); risk factor management; and exploration and solutions to problems affecting target behavior.

Information: Sources of Social or Peer Support or Adaptive Equipment
Five studies provided information on community resources, help from family and friends, patient support groups, modifying the home environment, and new adaptive equipment or techniques as one of the SM support components.

Remote Monitoring with Feedback and Action Plans
Seven studies addressed the components by phone calls, SMS, push notification, messaging devices, and online platforms. These include depressive symptoms, medication adherence, daily caloric intake, visual feedback on cardiovascular risk, verbal feedback on self-measurement, performance evaluation of targeted behavior, and discussion of future behavior formation.

Training/Rehearsal for Everyday Activities
Two studies consisted of exercise to improve physical function as one of the intervention sessions, two studies included the daily performance of targeted ADL tasks, and one study addressed daily calorie intake and exercise.

Clinical Review: Regular Follow-Up Reviews
In seven studies, the client's condition and SM by the healthcare expert were confirmed and reviewed weekly or after an intervention session, or specific time period during intervention.

Adherence Support
Seven of the ten studies included a logbook of participants' performance and symptoms, a reminder message about the task to be performed, and a motivational or positive encouragement message to enhance behavior adherence.

Psychological Support: Goal Setting, Action Planning, and Problem Solving Strategies
The psychological support component was used in nine studies. Although the studies did not include all detailed strategies for goal setting, action planning, and problem-solving, or did not clearly describe detailed strategies, they included at least one detailed strategy.

Social Support: Peer Support, Peer Mentoring, and Group Socialization
In two studies, social conversation and peer support among participants living in different regions were conducted through video conferencing.

Lifestyle Advice and Support: Practical Advice in Relation to Handling Life Stressors
Studies included support, discussion, or education for lifestyle changes related to physical activity, nutrition, relaxation, ADL performance, and healthy behavior, excluding one. Tables 5 and 6 show the outcomes of TH-SM interventions.   To assess the feasibility and acceptability of IVR as an adjunct to post-stroke depression follow-up care;

Outcome Measures
• Improved PHQ-9 scores from a median score of 11 (IQR 7-13) at baseline to a mean of 4 (IQR 1-7, p = 0.11) at follow-up • Reports from participants: good or excellent quality program, willingness recommend to others    # # +++: statistically significant effect; ++: greater improvement in intervention group than control but between group difference not significant; +: significant improvement in both groups but between group difference not reported or not significant; -: no reported change in the group(s) or between the groups; x: effect-related data not shown;ˆ: within-group improvement not significant; #: consequences, such as high compliance or positive feedback, TH: telehealth, SM: self-management, RCT: Randomized Controlled Trial, NRCT: Non-Randomized Controlled Trial, QOL: quality of life.
In studies using goal achievement as an outcome [20,25,26,28,29], only one study showed statistically significant improvement in group comparative studies [28], and one group of pre-and post-test studies reported an improvement in the TH-SM group, although this was not significant [26,29]. The results of goal achievement in various areas, such as function, participation, and environment, are shown [20,25]. Five studies measured the outcome of participation, and two studies reported a positive effect of improving participation [26,28].

Self-Efficacy
The efficacy of balance ability (n = 1), fall-efficacy (n = 1), and ADL performance (n = 1) were measured as self-efficacy outcomes. The TH-SM intervention had significant effects on all self-efficacy outcomes.

QOL
QOL was measured in three studies, and only one study reported a significant improvement [24].

Interventions
The major focus of TH-SM support in the selected studies was the relief of depressive symptoms, physical activity and participation, functional mobility, ADL, and obesity management in stroke survivors. PSD occurs in one-third of stroke survivors [30] and has a negative effect on the restriction of participation in the rehabilitation process, decline in physical, cognitive, and social functions, and the biological process of neuroplasticity [31]. However, according to previous studies, many patients with PSD are not treated, and accordingly, innovative strategies to identify and treat PSD are needed [32]. One study in this review used TH-SM support as one of the approaches to PSD and demonstrated positive results [27]. Participation restrictions are the difficulties encountered during the integration of stroke survivors into premorbid life roles [33]. The factors that determine the participation of stroke survivors are not only age and sex but also the survivor's functional/physical ability, independence in ADL, the severity of the stroke, and the onset of depression [34]. Traditionally, most SM interventions have primarily evaluated clinical outcomes, and SM support has focused on medical and emotional management skills, such as proper medication adherence and stress management in chronic conditions. However, outcomes related to function or participation are also important indicators and goals of SM support that help to improve the client's role management ability [35]. One study in this review addressed self-monitoring for obesity management [23]. Adherence to evidence-based approaches for obesity management in stroke patients is insufficient, but its effectiveness has been demonstrated in some limited healthcare interventions [36,37]. Recently smartphone-based self-monitoring intervention for obesity management has been documented as one of the useful mHealth interventions for managing stroke risk factors in stroke survivors with physical and cognitive challenges [38]. The studies included in this review provided TH-SM support, which focused on improving the role management ability and medical and emotional management skills of stroke survivors. It could be considered a necessary aspect for constructing SM support interventions with stroke survivors.

Types of TH Delivery
The nine SM support components of stroke survivors regrouped based on the PRISMS taxonomy for LTCs were provided through TH. Methods included messaging, telephone, video conferencing, and online platforms. In addition, TH intervention was provided either alone or in a combination of two types. Six out of ten studies used messaging as a tool to support, monitor, and provide customized information for target behavior. Champion et al. [39] reported that information on issues related to health behavior motivation and potential risk provided by messaging interventions could increase client knowledge and thus reduce health threats. This is consistent with the SM support interventions that can be loaded into the messaging delivery type identified in this review. Tele-communication can be adopted to improve the management of LTCs [40,41], and direct interactions, such as voice calls, between clients and healthcare professionals can contribute to building confidence with providing information and receiving immediate feedback [42]. The selected studies in this review used bidirectional telecommunication with phone calls or video conferencing, which were delivered for the purpose of monitoring psychological symptoms, medication adherence, and SM support education. Three studies provided information about stroke, cardiovascular risk and obesity management control monitoring, daily ratings on the client's own performance, and communication with healthcare professionals using a digital platform [23,24,29]. Evidence-based interventions, along with appropriate levels of guidance through a digital platform, can promote disease-specific health behavior changes and help effectively manage a variety of users and conditions [43]. The different types of TH delivery identified in this review are effective means for supporting TH-SM in stroke survivors. This may help with selecting an appropriate TH delivery type that can be equipped with a client-specific SM support strategy.

Outcomes and Effects of TH-SM Support Intervention
The outcomes included in this review were SM behavior, clinical outcome, self-efficacy, QOL, and TH acceptance. We could not find a consistent improvement in the detailed outcomes. Nevertheless, the outcomes that showed positive effects were the goal achievement of SM behavior, emotional state and mobility of clinical outcomes, and TH acceptance.
Goal achievement evaluations were used to facilitate goal setting and the program outcome evaluation tool. Goal setting is one of the core elements of SM support, especially in the rehabilitation setting, which is an important step in helping to promote community transitions among stroke survivors [44]. Mood and emotional disturbances are common symptoms of stroke survivors [45]. Three of the five studies that included the emotional state outcome reported the positive effects of TH-SM-supported interventions. Provided interventions were tailored suggestions for depressive symptoms, medication adherence, and obesity management monitoring, general SM education and exercise, and client-centered daily ADL tasks. Studies measuring mobility outcomes have provided physical exercise through video conferencing, and the results also showed positive effects on balance and walking ability. For stroke survivors who experience mobility restrictions after onset, appropriate physical activity is necessary for lifestyle changes and adaptation after acute care. Home telehealth can be effective for assessing the health care needs of stroke survivors and caregivers, as well as providing information and emotional support to them [46]. The TH acceptance addressed the participation rate and satisfaction survey of the provided SM-supported intervention and included the subjective opinions of the participants. Some subjective opinions on the technical deficiencies of TH are included in the provided intervention, but the TH acceptance of the participants was generally positive. Previous studies have reported remarkable acceptance and a positive attitude toward mobile-based intervention for stroke management [47]. However, the development and application of techniques that take into account the familiarity and comfort of stroke survivors and caregivers could be considered in the future.

Limitations
There have not been many TH-SM-supported studies in stroke survivors; therefore, only a few studies were included in this review. The focus of intervention varied for each method, so there was a limit to confirming consistent improvement in outcomes. In addition, the sample sizes were small, studies without a control group were included, and studies that included a control, waitlist, or usual care group did not provide inperson SM support intervention, which could not be compared with TM-SM-supported interventions. Stroke is an LTC, and it is a disease that requires changes in healthy behavior and continuous adaptation after onset. Therefore, it is necessary to confirm the long-term effects of SM intervention.

Conclusions
Although the TH-SM-supported intervention effects were not consistently found in all outcomes, this review discovered a positive effect on various SM-related outcomes. Therefore, we suggest TH-SM-supported intervention as an alternative method for SM support in stroke survivors. In addition, TH delivery types for SM support, the focus of interventions in stroke survivors, and the components or strategies used for effective implementation, which presents a variety of options for TH-SM support interventions and its applicability in stroke survivors. Based on these findings, various trials are needed to establish a consistent basis for the effectiveness of TH-SM intervention in stroke survivors in the future.  Institutional Review Board Statement: Not applicable.

Conflicts of Interest:
The authors declare no conflict of interest.