Feasibility of remotely supervised transcranial direct current stimulation and cognitive remediation: A systematic review

With technological advancements and an aging population, there is growing interest in delivering interventions at home. Transcranial Direct Current Stimulation (tDCS) and Cognitive Remediation (CR) as well as Cognitive Training (CT) have been widely studied, but mainly in laboratories or hospitals. Thus, the objectives of this review are to examine feasibility and the interventions components to support the domiciliary administration of tDCS and CR. We performed a systematic search of electronic databases, websites and reference lists of included articles from the first date available until October 31, 2018. Articles included had to meet the following criteria: original work published in English using human subjects, majority of tDCS or CR intervention administered remotely. A total of 39 studies were identified (16 tDCS, 23 CR/CT, 5 using both tDCS & CT). Four studies were single case studies and two were multiple case studies. The remaining 33 studies had a range of 9–135 participants. Five tDCS and nine CR/CT studies were double blind randomized controlled trials. Most studies focused on schizophrenia (8/39) and multiple sclerosis (8/39). Literature examined suggests the feasibility of delivering tDCS or CR/CT remotely with the support of information and communication technologies.


Introduction
Currently, 47 million people worldwide suffer from dementia, with nearly 10 million new cases each year, making it the 7th global leading cause of death [1]. Alzheimer's Dementia (AD) represents a growing health concern and contributing to 60-70% of dementia cases worldwide [2]. Given the growing prevalence rate of AD, preventative interventions and treatments that target individuals on a population level are crucial. PLOS  There are limited effective treatments available for AD, highlighting the necessity for preventative options. Recent research on preventative measures has focused on interventions that target brain neuroplasticity and cognitive reserve [3,4] due to observed maladaptive neuroplastic changes in various neuropsychiatric diseases [5]. Such changes are also visible in AD, whereby deficits in cognition, may be related to disruptions in the connections among neurons and neuronal networks [5]. Therefore, inhibiting these pathological changes and enhancing neuroplasticity may be beneficial for preventing or delaying the onset of AD and enhancing cognition [5]. Two interventions that have the potential of enhancing neuroplasticity and can be delivered remotely, offering a scalable preventative effect, are transcranial direct current stimulation (tDCS) and cognitive remediation (CR) [6,4].
CR has been shown in several trials that it could enhance cognition in healthy individuals and in patients with mild cognitive impairment (MCI), a clinical state that typically precedes AD [7][8][9]. Enhancing cognition in MCI is a promising strategy towards AD prevention [10]. Thus, efforts to optimize the efficacy of CR using synergistically acting interventions are highly needed. tDCS is an ideal intervention to be combined with CR because of its ease of use and portability to be delivered, for example, during CR, and because of its posited mechanism of action which would be complementary and possibly synergistic to CR as described above. Hence, in this review, we focused on the studies that combined CR with tDCS.
tDCS is a non-invasive brain stimulation that can be safely administered to awake outpatients and is thought to shift the resting membrane potential with a low intensity electrical current (e.g., 2 mA). This shift increases motor cortical excitability with anodal stimulation and suppresses it with cathodal stimulation [11]. However, this dichotomy is less consistent in non-motor cortical area [12]. tDCS has the potential to reduce symptoms of cognitive decline and enhance cognition and rehabilitation in neuropsychiatric diseases [4,13], including mild AD [14][15][16], through the modulation of neuronal activity and neuroplasticity [4,11]. Unlike other non-invasive brain stimulation devices, tDCS equipment is readily transportable, making it a viable population-level intervention, and an option for patients with AD who need remote assistance.
CR programs also offer a way of improving neurocognitive abilities by inducing functional changes within the brain [6]. In CR, patients engage in computerized cognitive exercises and, with the help of a therapist, are encouraged to utilize their metacognition in order to identify and modify their problem-solving techniques [17]. CR has been shown to improve cognition in schizophrenia [18], bipolar disorder [19,20], alcohol dependence [21], and major depression [6,22,64]. CR training programs are available online and therefore can be easily accessed from home.
tDCS is thought to alter cortical excitability resulting in more or less responsiveness to another intervention delivered concomitantly. As such, combining CR with tDCS could result in optimizing the effects of CR by priming the targeted cortical areas and, in turn, enhancing their response to CR. This is consistent with empirical evidence in healthy individuals [23].
Although a number of studies have examined the tolerability and efficacy of home-based, remotely-supervised tDCS and cognitive training in order to determine viability [24][25][26], no study to date has examined the combination of tDCS and CR delivered remotely as preventative measures for dementia. The current review intends to summarize existing literature on remote delivery of tDCS and CR/CT. Insight into the current research findings will allow for future determination of the potential usefulness of these two techniques to act as preventative treatment options for dementia on a population level.

Selection strategy
PubMed, Ovid, PsycINFO and CINAHL databases were searched focusing on studies from the first date available to October, 31 2018 for tDCS and CR/CT at-home studies. The literature search was divided into 2 categories: tDCS at-home studies. Titles and abstracts for the following Medical Subject Headings (MesH) terms and keywords were searched: (transcranial direct current stimulation or tDCS) in combination with "at-home" OR "home-based" OR "remotely supervised" OR "home treatment" OR "telemedicine" OR "self-administered." Cognitive training and cognitive remediation at-home studies. Titles and abstracts for the following MesH terms and keywords were searched: cognitive remediation or cognitive training in combination with "at-home" OR "home-based" OR "remotely supervised" OR "home treatment" OR "telemedicine" OR "self-administered." The rationale for searching for both CT and CR is the following: CT and CR are treatments that aim to improve cognition. In CT, participants typically engage with computer training exercises that use drill and practice to stimulate neuroplasticity. CT involves a set of tasks that focus on cognitive functions but CT does not include other aspects utilized in CR. CR is a behavioral intervention that utilizes a combination of drill and practice exercises, strategy contemplation, cognitive transfer as well as applying strategies to real-world situations. CR builds on CT with additional techniques such as monitoring and development of problem-solving strategies and considering how cognitive skills and strategies generalize to daily life. While both show efficacy for enhancing cognition, CR is more likely to result in changes in everyday functioning. CR, however, has not been utilized in studies targeting cognition remotely as often as CT. We found that by including studies utilizing CT, we would be able to get a better understanding of how these studies delivered this intervention remotely. This learning could then be applied and adapted to future studies looking at delivering CR remotely.
Selection criteria tDCS at-home studies. The following inclusion criteria were applied: (1) articles published in English; (2) original research with human participants, (3) home-based intervention. Excluded papers were: (1) articles reporting tDCS data from research settings, such as laboratories, hospitals, clinics, and research centers; (2) review, guideline and protocol papers without reporting original research; (3) articles that used animals as study subjects, (4) articles reporting do-it-yourself (DIY) tDCS use.
Cognitive training and cognitive remediation at-home studies. The following inclusion criteria were applied: (1) articles published in English; (2) original research with human participants; (3) home-based interventions. Excluded papers were: (1) articles reporting CR/CT data from research settings such as laboratories, nursing homes, hospitals, clinics, and research centers, (2) review, guideline and protocol papers without reporting original research.

Data extraction/collection
Data extraction is illustrated in Figs 1 and 2. After filtering for inclusion and exclusion criteria and eliminating duplicates, 11 home-based tDCS papers, 23 home-based CT or CR papers, and 5 studies that discussed both home-based CT and tDCS were included from this systematic multiple database search in the current review.  involving six participants and the other involving four participants. The remaining 33 published articles reported on a sample size range of 9-135 participants. Studies were published between the years of 1995-2018, and targeted multiple diagnoses including but not limited to: multiple sclerosis [24-26, 32-34, 51], Parkinson's disease [26,36,49], brain injury [50,58,60], and schizophrenia [35,43,45,52,54,55,57,61]. Only three papers [27][28][29] and one case study [30] were published with AD as the target population, with an additional study targeting mild vascular dementia [31]. Studies examined in this systematic search included both males and females with a noted age range of 7.77 (standard deviation: 1.62) to 94 years of age.
Participants. There was high variability in the sample size and age of participants across studies. The ages of participants ranged from 17 to 86. The number of participants in each study ranged from 1 to 45, with an average of 19.63 (SD 12.02) participants (see Table 1 for details).
Administration parameters. tDCS stimulation was administered for 20 minutes per session in all studies except three [30,35,39]. The current used for the stimulation ranged between 1 to 3mA, with 12 studies using either 1.5mA or 2mA. Administration time and the amount of current used for each study is listed in Table 1.
The number of sessions administered varied across studies with participants receiving an average of 15.17 (SD = 12.36) tDCS sessions, ranging between 4 to 60 sessions for studies lasting less than four months. Within longer running studies, one study involved 8 months of daily tDCS sessions [30], and another study followed a participant for three years receiving 1-2 daily treatments, with tDCS sessions still ongoing at the time of publication [35] (see Table 1 for details).
Training and safety measures. Although the majority of studies employed tDCS training at the baseline testing visit or during the first session at the research center (n = 12/16 studies) [22, 25, 26, 32-34, 36-39, 41, 42], the length, intensity and nature of training varied across research. In most cases training sessions consisted of an instructional video, hands-on tDCS coaching and monitoring by a study technician [26,32,34,36,39,41,42], troubleshooting techniques [26,34,39,41,42], as well as an assessment of the participant's ability to replicate the procedures competently at home [26,[38][39][40][41][42]. Training also incorporated safety assessments, including tolerability testing at the initial clinic visit [25, 26, 32-34, 36, 39, 42] (see Table 2 for details on training). Most studies implemented a number of safety measures in order to ensure safe and controlled use of the tDCS device. This included tDCS devices that were programmed to allow for a minimum of 12 hours between sessions [39], and devices that released a single session of stimulation after receiving a one-time unlock code [25, 26, 32-34, 36, 39, 42]. Additional safety measures included machines with password protected settings [40], anode and cathode sponges that had opposite male and female connections [40], as well as tDCS machines that measured and displayed contact quality prior to the start of stimulation [25,26,34,36,39,41,42].
Effectiveness of tDCS treatment. Overall, all studies reviewed reported significant improvement in physical or mental symptoms, such as a reduction in fatigue [25,33] and pain [22,25], improved pain management [41], improved anxiety and dizziness ratings [40], improvements in affect [25], motor function [36], rocking perception, [40], psychosocial functioning [35] and moderate improvements in consciousness [41]. Further significant improvements were noted in areas of cognition, including processing speed [25,31], visual recall [31], attention and response variability composites [32] as well as stabilization of cognitive decline for AD patients, with some improvement in memory function also noted [30].
Five studies addressed the combination of remotely delivered tDCS and cognitive training [25,26,32,33,36]. No additional training or supervision by study staff was discussed with the incorporation of cognitive training into the study protocol.
Participants. The age range and sample size varied, with participants ranging in age from 7 to 75. The number of participants in each study ranged from 1 to 135, with an average of number of 41.1 (SD = 35.55) participants.
Administration parameters. There was high variability in the administration of CR and CT. The duration of engagement, with one exception [50], ranged from 20 minutes per day to two hours per day, with a frequency of two to seven days per week.
Training and guidance. Participant training ranged in duration and intensity and consisted of no training and self-guided treatment (n = 4/24) [44,47,51,58], having one session of training in a group or alone (n = 6/24) [26,43,46,48,53,54], weekly training [50], observational training [27,28,59] or training on an as-needed basis [52]. Training generally involved printed instruction sheets and recommended strategies for daily activities, which were explained prior to starting the intervention [60], education surrounding the human brain, cognition and how cognition affects daily functioning [43,48], educational handbooks, worksheets and information about compensatory strategies [48], as well as the protocol for the computer software [56]. Only three studies discussed initial training, whereby training was conducted at the study center during the first session [48,54] for two hours [43] (see Table 4 for details on training and guidance for each study).
Feasibility of remote cognitive training and cognitive remediation. Most research reported remote, home-based cognitive remediation and cognitive training programs to be Remote tDCS and cognitive remediation   feasible, useful and well accepted for people with MS [24,51], schizophrenia [54,61], epilepsy [48], and those with brain damage [50,58]. Participants indicated that the training was beneficial, convenient [48], and enjoyable [48,54], and supported the benefit of remote CR using spousal caregivers for people with dementia [28]. Only one study noted that some participants felt the cognitive training instructions were confusing [56]. Remote tDCS and cognitive remediation

Discussion
The current literature suggests that both remotely administered tDCS and CR are feasible and possibly effective in targeting a number of cognitive functions in various patient populations. Proper training in equipment use and regular monitoring of procedures appear crucial for study compliance and feasibility. However, the field still needs to identify what components of both CR and tDCS administration are needed to be implemented in studying AD population or populations at risk for developing AD, e.g. patients with MCI.

tDCS
Based on the literature reviewed, the recommendations for future trials targeting cognitive decline in the AD population would include ensuring proper training and regular monitoring. This means dedicating adequate training time at the initial clinic visit or arranging in-home training in order to ensure hands-on coaching around tDCS application and proper use of the device. Explanation and demonstration of troubleshooting techniques could be helpful to include as part of the training procedure, including rectification techniques to manage pain and discomfort.
Considering the potential difficulty in tDCS self-administration with a predominantly elderly population, machine safety features and a customized headset could be useful. If a headset is not used, study partners or proxies can be utilized for tDCS application, with an anode and cathode that are clearly distinguishable to avoid interchangeability.
Video-monitoring as well as online and in-person check-ins could be beneficial ways for ensuring fidelity monitoring for a protocol targeting cognitive decline in the AD population. It is recommended that participants in future studies be overseen once every week, or every two weeks, through video conferencing or home visits by study staff. If tDCS stimulation is combined with an online cognitive training program, monitoring could also include remote control software. In addition, it is recommended that participants keep a log of sessions completed, either digitally or manually, in order to ensure protocol adherence.
The effectiveness of tDCS was dose-dependent, where more sessions (20 vs 10) and an increase in current (2.0mA vs 1.5mA) resulted in a greater reduction in fatigue [33,35]. This suggests that longer treatment periods and higher stimulation intensity are of greater benefit where at least 20 tDCS sessions with a current no lower than 2.0mA [24][25][26]62] may be optimal. Future trials might want to consider these settings to ensure maximum benefit.

Cognitive remediation
Ease of use may be a necessary factor to take into account for future trials in the AD population. Considering these patients already suffer from some cognitive decline and may be less experienced with computer software, a simpler computer program preinstalled by study staff and a longer training period might be preferred for tolerability. Additionally, having study staff engage in regular monitoring (i.e. regular phone calls, home visits or video conferencing) and using programs that ensure optimal cognitive challenge appear to be important components to consider for effectiveness and adherence.
Future trials may also want to consider utilizing exercises that target cognitive domains typically affected by AD related cognitive decline, such as memory, executive function, processing speed, attention, as well as reasoning and problem solving.
Moreover, continuously challenging cognition by finding an individualized optimal level of difficulty within the CR program used is an important aspect to consider for future trials targeting AD related cognitive decline. With the exception of four studies [29,48,54,58], all research reviewed had cognitive exercises that adapted in difficulty level based on the participant's performance. This self-adjusting feature may promote continued participant engagement and eliminate frustration brought about by seemingly unachievable difficulty parameters within the exercises.
Strategy contemplation [44,48,58] and cognitive transfer [28,48,58,60] are also important components to consider within the study design, whereby participants are asked to identify skills and strategies utilized in the exercises and apply them to real-world situations. This addition encourages cognitive activation and the utilization of adaptive problem-solving strategies in daily scenarios. To further cognitive transfer, some CR interventions encourage participants to seek out cognitively challenging activities in everyday life, and engage in activities that are cognitively stimulating outside of the program, however, this was not clearly visible in any of the studies reviewed. These add-on techniques could be particularly helpful in slowing cognitive decline and facilitating improvement in every day functioning and daily living within the AD population.

Limitations
A major limitation of this systematic review is that out of 39 studies reviewed, only four presented AD as the disease targeted. More specifically, within the cluster of at-home tDCS literature reviewed, only one case study used the intervention as a means to slow cognitive decline in an individual with early onset AD [30]. Similarly, within the group of at-home CR articles reviewed, one RCT targeted patients with AD [28], and one targeted individuals with Mild Cognitive Impairment (MCI) and AD [29]. A third article targeting AD utilized cognitive training rather than cognitive remediation [27]. This makes it difficult to make inferences and recommendations based on population specific difficulties encountered in past studies that could be improved upon in future studies.
Only six studies utilized discussion guided cognitive remediation (rather than cognitive training) as the observed intervention, with aspects of strategy awareness [61] and discussion [29,44,48,58] visible in five out of six CR studies reviewed, and the addition of cognitive transfer visible in 4 out of 6 [28,48,58,60]. As a result, it is difficult to evaluate the effectiveness and value of these add-on techniques within a trial targeting patients at-risk for developing AD. Furthermore, as different cognitive measures were used in each of the studies, it is difficult to determine if one method was more effective over another.

Conclusions
With millions of individuals being diagnosed with AD worldwide each year, and the accompanying maladaptive neuroplastic changes leading to worsening cognition, CR and tDCS are promising preventive interventions and treatments that can target individuals on a population level. Both CR and tDCS are effective interventions for improving cognition [6,[13][14][15][16][18][19][20][21] and have the benefit of being portable.
CR enhances frontal lobe activation and neuroplasticity [63] and it has been shown to improve cognition in depression [6,13,64]. CR's effectiveness relates to its inclusion of performance adapting software as well as strategy-based learning and bridging discussions, which typically occur in a group environment. However, with current technological advances, these components of CR can be achieved remotely.
tDCS modulates neuronal activity and enhances neuroplasticity [4,11] and has been shown to improve cognition in mild AD [14][15][16]. In other studies, participants must be present at a treatment centre in order to receive CR, tDCS or both from trained study staff. However, this can be costly and laborious for participants, and has the potential of being unfeasible for individuals with restricted mobility, vocational obligations, and lengthy travel times.
Future research may like to further consider additional measures that were not fully assessed throughout the literature examined, such as baseline ratings of reward (which may be positively associated with cognitive gains) and inclusion of a partner or caregiver. Investigating the feasibility of remotely delivering these interventions with other cohorts, such as individuals with mild AD, or those at risk of developing AD, who could benefit from at-home tDCS, CR or both would also be important considerations for prospective trials.