Remotely delivered information, training and support for informal caregivers of people with dementia

Summary of findings 1. Complex information, support and training interventions compared to usual treatment, waiting list or attention control informations for informal caregivers of people with dementia

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
Interventions involving training, support or both (with or without information) compared to usual treatment, waiting list or attention for informal caregivers of people with dementia
Participants or population: Informal caregivers of people with dementia Setting: Community Intervention: Any intervention including information, training or support Comparison: Usual treatment, waiting list or attention control
Outcomes	Risk with usual treatment, waiting list or attention	Risk with any intervention	Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
Caregiver burden assessed with: self‐reported scales ‐ lower score indicates less burden follow up: median 24 weeks	‐	SMD 0.06 lower (0.35 lower to 0.23 higher)	‐	597 (9 RCTs)	⊕⊕⊕⊝ MODERATE ¹	Any intervention probably does not affect caregiver burden. There are no important differences in effect estimates among type of interventions: SMD 0.12 higher for support or support with information SMD 0.14 lower for training or training with information SMD 0.17 lower for multicomponent interventions involving support and training
Caregiver mood assessed with: self‐reported scales ‐ lower score indicates fewer depressive symptoms follow up: median 24 weeks	‐	SMD 0.05 lower (0.22 lower to 0.12 higher)	‐	638 (8 RCTs)	⊕⊕⊕⊝ MODERATE ¹	Any intervention probably does not affect caregiver mood ‐ depressive symptoms. There are no important differences in effect estimates among type of interventions: SMD 0.06 lower for support or support with information SMD 0.02 higher for training or training with information SMD 0.06 lower for multicomponent interventions involving support and training
Caregiver health‐related quality of life (HRQoL) assessed with: self‐reported scales ‐ higher score indicates better QoL follow up: range 6 weeks to 24 weeks	‐	SMD 0.1 higher (0.13 lower to 0.32 higher)	‐	311 (2 RCTs)	⊕⊕⊕⊝ MODERATE ²	Any intervention probably does not affect caregiver health‐related quality of life. There are no important differences in effect estimates among type of interventions: SMD 0.12 lower for training or training with information SMD 0.15 higher for multicomponent interventions involving support and training
Dropouts for any reason assessed with: number of dropouts for any reason throughout the trial follow up: median 18 weeks	Study population		RR 1.15 (0.87 to 1.53)	661 (8 RCTs)	⊕⊕⊝⊝ LOW ^{3 4}	Any intervention probably results in little or no difference in dropouts for any reason. There are no important differences in effect estimates among type of interventions: RR 1.01 for support or support with information RR 1.42 for training or training with information RR 1.10 for multicomponent interventions involving support and training
	194 per 1000	223 per 1000 (169 to 296)	RR 1.15 (0.87 to 1.53)	661 (8 RCTs)	⊕⊕⊝⊝ LOW ^{3 4}
Use of health and social resources assessed with: Frequency of use throughout the trial follow up: mean 24 weeks	Study population		Rate ratio 1.05 (0.93 to 1.19)	4776 person‐weeks (1 RCT)	⊕⊕⊕⊝ MODERATE ⁵	Any intervention probably results in little to no difference in use of health and social resources.
	204 per 1000	214 per 1000 (190 to 243)	Rate ratio 1.05 (0.93 to 1.19)	4776 person‐weeks (1 RCT)	⊕⊕⊕⊝ MODERATE ⁵
Institutional care ‐ nursing home placement assessed with: Frequency of nursing home placement throughout the trial follow up: median 24 weeks	Study population		RR 0.59 (0.11 to 3.11)	34 (1 RCT)	⊕⊝⊝⊝ VERY LOW ^{6 7}	The evidence is very uncertain about the effect of any intervention on institutional care ‐ nursing home placement.
	188 per 1000	111 per 1000 (21 to 583)	RR 0.59 (0.11 to 3.11)	34 (1 RCT)	⊕⊝⊝⊝ VERY LOW ^{6 7}
*The risk in the intervention group (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: Confidence interval; HRQoL: Health‐related quality of life; QoL: Quality of life; RCT: Randomised controlled trial; RR: Risk ratio; SMD: Standardised mean difference
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect
¹Downgraded for study limitations: Presence of high risk for performance and detection bias due to lack of blinding for participants, personnel, and the use of self‐reported outcome scales by unblinded participants. Also, presence of attrition bias. ²Downgraded for study limitations: Presence of high risk for performance bias. ³Downgraded for study limitations: Presence of high risk for performance and detection bias. Lack of blinding for participants might affect the overall rate of dropouts. ⁴Downgraded for imprecision: Few events to get precise estimates (< 300 overall). 95% CI includes a negligible effect and important benefit/harm. ⁵ Downgraded for imprecision: Rate ratio estimated with 1 study only including 250 participants and 4776 person‐weeks of follow up. ⁶Downgraded for study limitations: Unclear risk for selection bias. High risk for performance and detection bias. ⁷Downgraded for imprecision: Only 1 study with very few events (5 overall).

Summary of findings 2. Complex information, support and training interventions compared to information provision only for informal caregivers of people with dementia

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
Interventions involving training, support or both (with or without information) compared to control information alone for informal caregivers of people with dementia
Participants or population: Informal caregivers of people with dementia Setting: Community Intervention: Any intervention including information, training or support Comparison: Information
Outcomes	Risk with control information	Risk with any intervention	Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
Caregiver burden assessed with: self‐reported scales ‐ lower score indicates less burden follow up: median 20 weeks	‐	SMD 0.24 lower (0.51 lower to 0.04 higher)	‐	650 (9 RCTs)	⊕⊕⊝⊝ LOW ^{1 2}	Any intervention may result in a slight reduction in caregiver burden. There are no important differences in effect estimates among type of interventions: SMD 0.03 lower for support or support with information SMD 0.31 lower for training or training with information SMD 0.11 lower for multicomponent interventions involving support and training
Caregiver mood assessed with: self‐reported scales ‐ lower score indicates fewer depressive symptoms follow up: median 20 weeks	‐	SMD 0.25 lower (0.43 lower to 0.06 lower)	‐	1100 (11 RCTs)	⊕⊕⊕⊝ MODERATE ¹	Any intervention probably results in a slight improvement in caregiver mood ‐ depressive symptoms. There are no important differences in effect estimates among type of interventions: SMD 0.28 lower for support or support with information SMD 0.31 lower for training or training with information SMD 0.04 higher for multicomponent interventions involving support and training
Caregiver health‐related quality of life (HRQoL) assessed with: self‐reported scales ‐ higher score indicates better QoL follow up: range 12 weeks to 24 weeks	‐	SMD 0.03 lower (0.28 lower to 0.21 higher)	‐	257 (2 RCTs)	⊕⊕⊝⊝ LOW ^{1 3}	Any intervention may result in little or no difference in caregiver health‐related quality of life.
Dropouts for any reason assessed with: number of dropouts for any reason throughout the trial follow up: median 16 weeks	Study population		RR 1.51 (1.04 to 2.20)	1266 (12 RCTs)	⊕⊕⊕⊝ MODERATE ¹	Any intervention probably results in an increase in dropouts for any reason. There are no important differences in effect estimates among type of interventions: RR 1.10 for support or support with information RR 1.51 for training or training with information RR 1.25 for multicomponent interventions involving support and training
	130 per 1000	196 per 1000 (135 to 286)	RR 1.51 (1.04 to 2.20)	1266 (12 RCTs)	⊕⊕⊕⊝ MODERATE ¹
Use of health and social resources ‐ not reported	‐	‐	‐	‐	‐	No included study reported use of health and social resources.
Institutional care ‐ nursing home placement assessed with: frequency of nursing home placement throughout the trial follow up: mean 12 weeks	Study population		RR 2.67 (0.12 to 60.93)	32 (1 RCT)	⊕⊝⊝⊝ VERY LOW ^{4 5}	The evidence is very uncertain about the effect of any intervention on institutional care ‐ nursing home placement.
	0 per 1000	0 per 1000 (0 to 0)	RR 2.67 (0.12 to 60.93)	32 (1 RCT)	⊕⊝⊝⊝ VERY LOW ^{4 5}
*The risk in the intervention group (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: Confidence interval; HRQoL: Health‐related quality of life; QoL: Quality of life; RCT: Randomised controlled trial; RR: Risk ratio; SMD: Standardised mean difference
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect
¹Downgraded for study limitations: Presence of high risk for performance and detection bias due to the lack of blinding for participants, personnel, and the use of self‐reported outcome scales by unblinded participants. Also, presence of attrition bias. ²Downgraded for inconsistency: Unexplained important heterogeneity (I² > 60% or not overlapping 95% confidence intervals). ³Downgraded for imprecision: Few participants to get precise estimates (< 300 overall). ⁴Downgraded for study limitations: Presence of high risk for performance and detection bias due to lack of blinding for participants, personnel, and the use of self‐reported outcome scales by unblinded participants. ⁵Downgraded for imprecision: Few events to get precise estimates (< 300 overall). 95% CI includes a negligible effect and important benefit/harm.

Background

Description of the condition

It is estimated that worldwide, about 50 million people have dementia, and there are nearly 10 million new cases every year (WHO 2020). Many people with dementia are being cared for at home by an unpaid informal caregiver, who is usually a female family member, often a spouse or daughter. It is estimated that family caregivers spend five to 20 hours per day and an average of 60 hours per week caring for the person living with dementia (Marziali 2006).

As a consequence of their role as caregiver of a person with a progressive and irreversible disease, the informal caregiver may experience a range of physical, emotional, financial and social harms (Collins 2011). This situation has been widely described in the literature and is known as caregiver burden (Dillehay 1990; George 1986). Caregiver burden is associated with characteristics of the caregiver, such as gender, and characteristics of the person with dementia, such as behavioural problems, stage of dementia and the presence of neuropsychiatric disturbances (Haley 2004). It is a strong predictor of a move from home care to residential care for people with dementia (Eska 2013).

Description of the intervention

A variety of psychosocial interventions have been suggested to prevent or to reduce the negative consequences of dementia care on informal caregivers (Cooke 2001; Etters 2008; Pusey 2001; Selwood 2007). The interventions are generally heterogeneous regarding (i) their theoretical framework, (ii) the focus and the type of interventions reported (e.g. behavioural intervention, stress management, education about the disease), and (iii) the administration format (group or individual, generic or tailored to specific needs, telephone‐ or Internet‐based).

Interventions that are delivered remotely using written materials, the telephone or Internet may be particularly suitable for caregivers of people with dementia who can struggle to access face‐to‐face services because of their own health problems, poor access to transport, or difficulties finding substitute care. At the time of writing, measures used by many countries to contain the COVID‐19 pandemic require that most people, including people with dementia and their family carers, are confined to their homes, isolating themselves from direct contact with others. In such circumstances, there is no alternative to remote delivery of interventions.

There are also, however, significant barriers to remote delivery of interventions, particularly for older caregivers. They may have sensory impairments or other disabilities that make use of the necessary technologies difficult. They may also lack confidence with the technologies involved and may particularly value direct contact with others, being less familiar than many younger people with the experience of finding support in online communities.

This review assesses the efficacy and acceptability of remotely delivered interventions that aim to offer information, training and support, or a mixture of these, to informal carers of people with dementia.

Information

We considered information interventions to have two key elements: (i) they involve the provision of standardised information whose contents are determined by the professional; and (ii) the participant has a passive role (i.e. there is no interaction with the professional or active practice of skills). The intervention can address various aspects: issues related to the person with dementia (e.g. signs and symptoms, natural history, treatments and care alternatives), their care (e.g. activities of daily living, nutrition, communication with the person with dementia), the difficulties derived from their care (behavioural problems, role incompatibility, use of drugs, sleep disturbances), or resources available in the community (Gallagher‐Thompson 2010). The format of implementation can vary: leaflets, manuals, lectures or audio‐visual presentations. The nature of these interventions makes them relatively easy to deliver remotely.

Training

The key element of training interventions is to provide caregivers with the practical skills to manage the burden of care. Caregivers will play an active role in the program (e.g. through supervised practice, or role playing). The training may relate to the care of the person with dementia (e.g. recognising behavioural triggers, communicating more effectively, tailoring tasks to patients' capabilities) or to management of the caregiver's psychological stress. Training can be classified into the following categories: (i) meeting patients' basic care needs, (ii) managing behavioural problems, and (iii) managing and coping with caregiver stress (Losada Baltar 2004; Martin‐Carrasco 2014). Delivery of these elements remotely is likely to be more challenging than the simple provision of information.

Support

The key element of support interventions is the participants' active role in discussing and sharing feelings, problems or issues related to care with other caregivers and professionals (Winter 2006). This category could include two types of support: (i) peer support (e.g. participation in a caregivers' group that is not professionally facilitated), and (ii) professional support (e.g. receiving counselling from a health professional) (Mittelman 2004). Peer support groups, in particular, are usually face‐to‐face, but as technologies develop, it is increasingly possible for these to take place online.

How the intervention might work

Interventions for caregivers of people with dementia may work in a variety of ways depending on the content. Broadly speaking, they are intended to improve caregivers' knowledge about the disease and its care (Ducharme 2011; Liddle 2012); to increase their sense of competence (Laakkonen 2012) and their ability to cope with difficult situations (e.g. behavioural disturbances, communication problems) (Cheng 2012); and to relieve feelings of isolation, as well as to allow caregivers to attend to their own emotional and physical needs (Roth 2009).

Why it is important to do this review

Throughout the world, the majority of care for people with dementia is provided by family members. Anything that can be done to ease the burden of this care will be good for the caregivers and for the people they care for. It may also ease the burden on health and social care systems, for example by improving caregivers' health and by delaying institutional care for the person with dementia. All health and care systems are struggling with the magnitude of the problem presented by dementia. Remote delivery of interventions to support caregivers, if acceptable and effective, offers the opportunity to reach more people than conventional services, potentially at lower cost.

Development of remotely delivered interventions acquired a new urgency in the context of the COVID‐19 pandemic. The strain on family caregivers who are isolated with a person with dementia is hugely magnified. Not only are they cut off from their usual community and professional supports, but so is the person with dementia, whose behaviour may be more challenging as a result. Because of their age and vulnerability, these groups may experience more prolonged periods of isolation than other members of the community. Identification of effective interventions for caregivers that can be delivered remotely is needed more than ever.

This review aims to collate the best evidence about the efficacy and acceptability of these interventions to facilitate decision‐making by those designing, delivering and receiving services.

Objectives

To assess the efficacy and acceptability of remotely delivered interventions aiming to reduce the burden and improve the mood and quality of life (QoL) of informal caregivers of people with dementia.

Methods

Criteria for considering studies for this review

Types of studies

We included randomised controlled trials (RCTs) in which caregivers were randomly assigned to remotely delivered psychoeducational interventions that included information with support or training, or to a control condition. The control condition could be simple provision of information or waiting list, usual care or a non‐specific intervention serving as an attention control.

We excluded all non‐randomised studies and also RCTs with a cross‐over design, due to the high risk of carryover effects.

Types of participants

Eligible trials included informal caregivers of people with dementia receiving care at home. The dementia of the cared‐for person could be of any type. We accepted clinical diagnoses or diagnoses reached using formal diagnostic criteria such as the Diagnostic and Statistical Manual of Mental Disorders (DSM), the International Classification of Diseases (ICD), the National Institute on Aging ‐ Alzheimer's Association (NIA‐AA) or the Alzheimer's criteria of the National Institute of Neurological and Communicative Disorders and the Alzheimer's Disease and Related Disorders Association (NINCDS‐ADRDA). Caregivers had to be unpaid adults (relatives, or members of the person's community; 18 years of age or older), of either sex, and of any ethnic or geographical origin. Studies that included caregivers of people with mixed diagnoses were included only if data on caregivers of people with dementia could be extracted separately.

Types of interventions

We included trials that assessed remote delivery of structured interventions involving training, support, information provision, or combinations of these elements, and that only targeted informal caregivers. We considered an intervention to be structured if there was a clear description of a standard set of components and how they were implemented. They could be delivered using paper, the telephone, the Internet or a mixture of these, but did not involve any face‐to‐face contact with professionals or other caregivers. These interventions had to be compared with a control condition, which could be provision of information alone or a waiting list, usual care or any non‐specific intervention serving as an attention control. Participants in waiting list groups were offered the experimental interventions once the study finished. Participants in attention control groups received a similar amount of contact with professionals to those in the experimental group, but the content of the contact involved no specific information, training or support related to dementia.

The following interventions were excluded:

interventions aimed at healthcare professionals;
interventions aimed at the people with dementia;
respite care interventions;
interventions for caregivers that are individually tailored and whose components are not well defined; and
interventions for caregivers that are predominantly psychotherapy, including interventions based entirely on the cognitive behavioural model (CBT). However, interventions that included elements of CBT as part of a larger, multimodal intervention were eligible for inclusion.

We used the following criteria to identify information, support and training. Information interventions include two key elements: (i) they provide standardised information, and (ii) the caregiver plays a passive role. Support interventions promote caregiver interaction with other people (professionals or peers). Training interventions are structured interventions intended to train caregivers in practical skills in order to prevent or alleviate the negative consequences of care giving.

We used the following operational criteria to categorise the interventions:

Intervention

Definition

Key elements

Information

Structured programmes with standardised information about issues related to care of people with dementia and caregiver burden.

The contents of the intervention can include issues related to:

Manifestation of dementia (e.g. signs and symptoms, natural history, treatments and care alternatives).
Care of the people with dementia (e.g. activities of daily living, nutrition, communication with the person with dementia).
Difficulties associated with their care (e.g. managing behavioural problems, role incompatibilities, use of drugs, sleep disturbances), or resources available in the community.

The intervention involves the provision of standardised information whose contents are determined by a professional in charge of the dementia.

The participant has a passive role, and there is no interaction with the professional or active learning and practice of appropriate skills.

Support

Programmes that allow the caregiver to talk, discuss or share information about their caregiver issues.

There are two types of support:

Peer support (e.g. participating in a caregivers' group that is not professionally facilitated).
Professional support (e.g. having counselling from a health professional).

The participant has an interactive role talking, discussing or sharing their feelings, problems or contents related to the care provided with other caregivers or professionals.

Training

Structured interventions intended to empower caregivers with practical skills to manage the issues related to care.

The skills to be acquired may relate to the care of the person with dementia, or to the management of the caregiver's psychological stress, according to the following categories:

Fulfilling basic care needs.
Management of behavioural problems.
Stress management and coping style.

Caregivers play an active role (e.g. through supervised practice, role playing, etc.) in the intervention provided.

We compared (1) all interventions involving training, support or both, with or without information provision, with usual treatment, waiting list or attention control and (2) all interventions involving training, support or both, with or without information provision, with provision of information alone. We used the Template for Intervention Description and Replication (TIDieR) checklist to describe the interventions tested in each included trial, to obtain a detailed account of the assessed interventions, and to improve comparability among studies (Hoffmann 2014).

Types of outcome measures

We focused on outcomes that have been identified as relevant to evaluate effectiveness of psychosocial interventions in dementia care (Moniz‐Cook 2008).

Primary outcomes

Caregiver burden, measured by validated questionnaires such as the Zarit Burden Interview (ZBI) or the Sense of Competence Questionnaire (SCQ 27).
Caregiver mood or psychosocial well‐being, measured by validated questionnaires such as the Center for Epidemiologic Studies Depression Scale (CES‐D), the Hospital Anxiety and Depression Scale (HADS), or the General Health Questionnaire (GHQ).
Caregiver health‐related quality of life (HRQoL), measured by the EuroQoL Group's instrument (EUROQOL), the 36‐Item Short Form Health Survey (SF‐36) or the World Health Organization (WHO) Quality of Life assessment tool (WHOQOL).
Acceptability of the intervention: we used dropouts for any reason as a surrogate for the acceptability of interventions.

Secondary outcomes

Caregiver knowledge and skills.
Use of health or social resources.
Admission of the person with dementia to institutional care.
QoL of the person with dementia, measured by validated questionnaires such as the specific Quality of life in Alzheimer's disease (QOL‐AD) instrument, the Dementia Quality of Life (DQOL) instrument, or the generic EQ‐5D instrument.

Search methods for identification of studies

Electronic searches

We searched ALOIS, the Specialised Register of the Cochrane Dementia and Cognitive Improvement Group (www.medicine.ox.ac.uk/alois) on 10 April 2020. We used the following terms: (caregiv* or carer*) and (support* or information* or program* or training* or psychoeducation* or psycho‐education* or skill* or education* or tele* or video* or computer* or internet* or online or stress*). In addition, when trials are added to the ALOIS database, they are assigned a 'study aim.' One of the possible study aims is "caregiver focussed." We searched the whole database using this aim filter.

ALOIS is maintained by the Information Specialists of the Cochrane Dementia and Cognitive Improvement Group (CDCIG), and contains studies that fall within the areas of dementia prevention, dementia treatment and management, and cognitive enhancement in healthy elderly populations. The studies are identified through:

Monthly searches of a number of major healthcare databases: MEDLINE, Embase, CINAHL, PsycInfo and LILACS;
Monthly searches of a number of trial registers: ISRCTN; UMIN (Japan's Trial Register); the World Health Organization International Clinical Trials Registry Platform (WHO ICTRP) portal (which covers ClinicalTrials.gov; ISRCTN; the Chinese Clinical Trials Register; the German Clinical Trials Register; the Iranian Registry of Clinical Trials and the Netherlands National Trials Register, plus others);
Quarterly search of the Cochrane Central Register of Controlled Trials (CENTRAL);
Six‐monthly searches of a number of grey literature source: ISI Web of Science Core Collection.

To view a list of all sources searched for ALOIS, see About ALOIS on the ALOIS website (www.medicine.ox.ac.uk/alois).

We also ran additional searches in MEDLINE, Embase, PsycINFO, CINAHL, and CENTRAL. We searched two international trials registries, ClinicalTrials.gov at the US National Institutes of Health, and the WHO ICTRP, to ensure that the search was as comprehensive and as up‐to‐date as possible. We have presented the search strategies used and the number of results retrieved in Appendix 1.

The most recent search for this review was done on 10 April 2020. Before that date, we had run searches in December 2015, July 2016, May 2017, May 2018 and May 2019.

Searching other resources

We complemented this search with manual searches of the bibliographies of relevant review papers and published trials to locate additional relevant studies.

Data collection and analysis

Selection of studies

Pairs of review authors (two of EGF, JRR, JB) independently examined the titles and abstracts of citations obtained by the searches to identify studies that may meet the inclusion criteria. We retrieved the full‐texts of all potentially eligible articles and examined these against the inclusion and exclusion criteria, again independently and in duplicate. Studies excluded and reasons for exclusion are reported in the Characteristics of excluded studies table.

Data extraction and management

Pairs of review authors (two of EGF, JRR, BS, JB) independently extracted data using a standardised data collection form. In case of disagreement, another review author (IS) was consulted and consensus sought. For each study, extracted data included: trial registry identification, funding and potential conflicts of interest, main methodological characteristics, results. We described the key characteristics of each trial in tables, paying special attention to the characteristics of the interventions assessed. We described the interventions according to the TiDieR checklist (Hoffmann 2014).

Assessment of risk of bias in included studies

We assessed the risk of bias in included studies by following the guidance of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). Pairs of review authors (two of EGF, JRR, JB) extracted the appropriate information and independently rated the risk of bias for each study and outcome. In case of disagreement, another review author (IS) was consulted and consensus sought. We assessed the following sources of bias: selection bias (including random sequence generation and allocation concealment); performance bias (blinding of participants and personnel); detection bias (blinding of outcome assessments); attrition bias (incomplete outcome data); and reporting bias (selective reporting).

Using 'Risk of bias' tables, we rated the risk of bias in each domain as either "low risk," "unclear risk" or "high risk" and provided an explanation for each rating. We considered trials with inadequate randomisation and lack of blinding (for outcome assessment) as being at overall "high risk of bias."

Measures of treatment effect

We used the risk ratio (RR) with its 95% confidence interval (CI) as the measure of treatment effect for dichotomous data and the mean difference (MD) with its 95% CI as the measure of treatment effect for continuous data. We used the SMD with its 95% CI for continuous outcomes only if similar outcome constructs were measured by different rating scales.

Unit of analysis issues

The unit of analysis was the individual allocated to a treatment group. We did not identify any eligible cluster‐randomised trials. If a trial reported outcomes at more than one time point (e.g. weekly or monthly outcome measures; end‐of‐trial and follow‐up outcome measures), we extracted the measure closest in time to the end of treatment and used this as a measure of the acute effect of the intervention. If a trial included more than one control or experimental group, we followed the guidance of the updated Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2020). When managing several eligible experimental interventions versus a unique control, we split the control sample to avoid over‐counting participants. Where there was more than one control group, we combined these if it was conceptually reasonable to do so.

Dealing with missing data

We used intention‐to‐treat (ITT) data where this was reported. If an included study imputed missing data, we reported the data imputation method.

Assessment of heterogeneity

We considered heterogeneity in study characteristics to decide if combining the results would be clinically meaningful. If we judged the studies to be too heterogeneous, then we did not conduct meta‐analyses, but reported individual results. We assessed between‐studies heterogeneity in meta‐analyses using the I² statistic, complemented with the examination of overlapping 95% CI.

Assessment of reporting biases

We assessed small study bias using funnel plots where there were 10 or more studies in a comparison.

In the bibliographic searches, if we found any RCT included in any clinical trial register that we considered could have been completed, but for which we had not been able to locate any results, we contacted the authors to request information about the status of the trial, and the trial results, if it had been completed.

Data synthesis

If we considered trials to be sufficiently similar that it was appropriate to pool data, we conducted meta‐analyses using a random‐effects model for two main comparisons:

a comparison of all interventions involving information, training or support versus usual treatment, waiting list or attention control;
a comparison of all interventions involving training or support (with or without information) versus information alone.

Where there were sufficient data, we stratified the analyses by type of intervention.

We conducted all analyses in Review Manager version 5.3 (Review Manager 2014).

Subgroup analysis and investigation of heterogeneity

Depending on availability of data, we intended to conduct subgroup analyses according to:

the intensity/length of the intervention provided;
the mode of delivery of the intervention: telephone‐only versus Internet with or without telephone;
the characteristics of participants; and
the outcome scales used to score each construct of interest.

Sensitivity analysis

If feasible, we intended to perform sensitivity analyses to assess the influence of studies at higher risk of bias, and the impact of attrition on the robustness of the pooled results. However, the small number of included studies by group comparison precluded reliable sensitivity analyses. Another form of sensitivity analysis, the leave‐one‐study method, did not identify any specific study as highly influential on the combined results.

'Summary of findings' tables

We summarised the review findings in a 'Summary of findings' table using the online GRADEpro GDT application (GRADEpro GDT). In the table, we reported the estimated treatment effects for the review's primary outcomes and for resources use and institutionalisation. For each effect estimate, we used the GRADE approach (Schünemann 2013) to rate our confidence that the estimate is correct (high, moderate, low or very low certainty). GRADE ratings take account of study limitations, imprecision of effect estimates, inconsistency among studies, indirectness of evidence and publication bias. When assessing imprecision, we considered that an absolute value of SMD 0.50 probably represented an important between‐group difference. We considered inconsistency to be substantial enough to downgrade our confidence in the effect estimate if we found I² to be greater than 60%.

Results

Description of studies

Results of the search

After a first assessment by the CDCIG information specialist, we assessed the remaining 2,010 potentially relevant records. Of these, 264 records remained after first assessment by the author team, and after more detailed screening of titles and abstracts, we retrieved 94 full‐text articles for assessment. We excluded 52 articles, recording the reasons for exclusion. We included 26 studies (39 references) in the qualitative synthesis and 25 studies in the quantitative synthesis. Three studies are awaiting classification because they may be eligible, but there is not enough information in their registry protocols to make an eligibility decision (see Studies awaiting classification). Another 13 studies are ongoing studies that have not yet reported results (see Ongoing studies). Figure 1 presents the flow diagram of the study selection process.

Figure 1

Study flow diagram.

Included studies

We identified 26 studies with 2367 participants for inclusion in the review (see Characteristics of included studies).

Setting and participants

Included studies were conducted or published over 25 years (from 1995 to 2020). Most studies were from the US (15 studies, 58%). China and the Netherlands contributed three studies each, France two studies, and Canada, Spain and the UK one study each. Studies had a median sample size of 67 participants (interquartile range (IQR) 49 to 110) and a median duration of 16 weeks (IQR 12 to 24). Only six studies (23%) provided data for a continuation follow‐up after the end of the intervention (median of 24 weeks).

Most caregivers were female (72%) with a median age of 63 years (range 51 to 72 years) . Care recipients were mainly diagnosed with dementia due to Alzheimer's disease (83%), had a median age of 78 years (range 62 to 81 years) and had been treated for dementia for a median of 38.5 months (range 30 to 53 months).

Interventions

We examined the descriptions of the experimental interventions in the included studies to look for elements meeting our definitions of information, training and support. Almost all of the experimental interventions specified provision of information to caregivers as a component of a more complex intervention. We considered this element was not present in only three studies: Davis 2004 and Dowling 2014, which we considered to be relatively pure training interventions, and Winter 2007, which we considered to be a relatively pure support intervention. However, even in these cases, it was possible that some provision of information did occur during the course of delivery of the other elements. Therefore, we constructed three subgroups of experimental intervention based on the predominant components: training with or without information, support with or without information, and interventions including both support and training elements. We considered these groups to be exploratory, recognising that all the interventions were complex in nature and that there is inevitably some overlap among the categories.

We classified the experimental intervention in 12 studies as training or training with information (Au 2015; Au 2019; Blom 2015; Czaja 2013; Davis 2004; Dowling 2014; Gant 2007; Kajiyama 2013; Martindale‐Adams 2013; Metcalfe 2019; NCT00056316; NCT03417219), in eight studies as support or support with information (Brennan 1995; Hattink 2015; Hayden 2012; Kwok 2013; Mahoney 2003; Nunez‐Naveira 2016; Torkamani 2014; Winter 2007), and in six studies as involving both training and support with information (Cristancho‐Lacroix 2015; Duggleby 2018; Gustafson 2019; Huis in het Veld 2020; Tremont 2008; Tremont 2015). Fourteen studies delivered the remote intervention by telephone and 12 using the Internet.

Eight studies included usual treatment (Brennan 1995; Cristancho‐Lacroix 2015; Mahoney 2003; Nunez‐Naveira 2016; Torkamani 2014; Tremont 2008) or waiting list (Hattink 2015; Metcalfe 2019) as the control condition. Four studies provided some form of attention control during the treatment period, with or without socially supportive telephone calls (Davis 2004; Dowling 2014; Hayden 2012; Tremont 2015). Eight studies provided specific information on dementia and related problems (Blom 2015; Duggleby 2018; Gustafson 2019; Kajiyama 2013; Kwok 2013; Martindale‐Adams 2013; NCT03417219; Winter 2007) in the control condition. Six further studies provided both information and some form of control attention (Au 2015; Au 2019; Czaja 2013; Gant 2007; Huis in het Veld 2020; NCT00056316).

The TIDieR checklist for included studies is reported in Additional tables (Table 1; Table 2; Table 3; Table 4; Table 5; Table 6; Table 7; Table 8; Table 9; Table 10; Table 11; Table 12; Table 13; Table 14; Table 15; Table 16; Table 17; Table 18; Table 19; Table 20; Table 21; Table 22; Table 23; Table 24; Table 25; Table 26 ).

Table 1. TIDieR ‐ Au 2015

Study	Au 2015
TIDieR item	Experimental intervention	Control intervention
BRIEF NAME	Psychoeducation with behavioral activation (PsyED‐BA).	Psychoeducation only (PsyED).
WHY	Based on the Gallagher‐Thompson’s program Coping with Cargiving. The aim is to enhance the use of coping skills via behavioral activation (BA), since it would be more acceptable and helpful than change dysfunctional thoughts, overcome barriers to treatment access and train paraprofessionals to deliver some modules of the program.	General discussion about the psychoeducation program and information packet.
WHAT materials	Caregivers were given a printed copy of the psychoeducation program (adapted from the Chinese version of the Coping with Caregiving manual) together with an information packet with fact sheets concerning local organizations, community resources, and social and mental issues related to dementia. 1.‐ Patient‐Change Workshop, Procedures Manual, Progress Notes, Problem Behavior Tracking forms, written cueing systems for repetitive verbal behavior, audio/videotape or tactile diversion program for repetitive vocal and physical behaviours, and verbal and physical prompting procedures for continence and self‐feeding problems, Caregiver Self‐Change Recording Form. 2.‐ Self‐change Workshop, Problem Behavior Tracking forms, Caregiver‐self change recording form.	Caregivers were given a printed copy of the psychoeducation program (adapted from the Chinese version of the Coping with Caregiving manual) together with an information packet with fact sheets concerning local organizations, community resources, and social and mental issues related to dementia.
PROCEDURES	Participants were contacted by telephone within the same week of the baseline assessment to start the psychoeducation program. There were a total of four weekly intervention telephone calls, each lasting about 30 minutes. ‐ In the first week, all participants were taught about the symptoms and associated behavioral changes of dementia and the possible effects on the caregivers. Participants were also invited to share their caregiving experiences. ‐ In the second week, participants learned about the physical, social, and psychological consequences of stress and being aware of their own stress. They were invited to talk about their own stress. ‐ In the third week, they learned about the principles of identifying and scheduling pleasant events for themselves. ‐ In the fourth week, participants learned about communication: communicating their own needs to seek support from their family members. They also learned about the characteristics of various different types of communications: passive, aggressive, and assertive. Participants had eight biweekly telephone sessions over the following 4 months. Each session consisted of a telephone call lasting 15 to 20 minutes. The first four sessions focused on pleasant event scheduling and the other four sessions on effective communication. The tasks involved in each of the four sessions were as follows: 1. Activity monitoring: how is the participant spending time/ communicating now? 2. Activity scheduling: schedule pleasant event/effective communications 3. Reinforcing or modifying the pleasant event and communication based on feedback or self‐evaluation 4. Activity rescheduling/revision based on changes after modification. Finally, social work services were available upon requests in outpatient departments in which the care recipients received their regular follow‐up.	Participants were contacted by telephone within the same week of the baseline assessment to start the psychoeducation program. There were a total of four weekly intervention telephone calls, each lasting about 30 minutes. Participants had eight biweekly telephone sessions over the following 4 months. Each session consisted of a telephone call lasting 15 to 20 minutes: ‐ Participants were asked to go through the materials of the psychoeducation program and the information package. For each of the telephone session, the participants were asked to select their own topics for general discussion. If the caregiver selected pleasant event scheduling or communications, general discussion would be carried out without any BA procedures. Finally, social work services were available upon requests in outpatient departments in which the care recipients received their regular follow‐up.
WHO provided	Five paraprofessionals recruited trained to administer BA procedures. They were between 55 and 60 years old and had previously completed post‐secondary school training in areas related to human services (i.e., nursing or management). They had subsequently completed a 42‐hour course on Introduction to Psychology and had then received 20 hours of group training led by a social worker and a clinical psychologist (the principal investigator of the study) on BA in the context of the Copy with Caregiving program. Each paraprofessional worker was tested on a mock case before delivering the program to actual caregivers. Ongoing weekly supervision was provided by the clinical psychologist and social worker.
HOW delivered	Individually over the telephone.
WHERE occurred	At home.
WHEN and HOW MUCH	Twelve sessions: sessions 1 to 4 were delivered weekly, sessions 5 to 12 were delivered biweekly. Each session consisted of a telephone call lasting 15 to 20 minutes.
TAILORING	Interventions were comprehensively designed, not tailored to cover individual or unmeet needs.
MODIFICATIONS	None described.
HOW WELL planned	No described.
HOW WELL actual	No described.

Table 2. TIDier ‐ Au 2019

Study	Au 2019
TIDieR item	Experimental intervention	Control intervention
BRIEF NAME	Psychoeducation with behavioral activation (TBA).	General Monitoring (TMG).
WHY	Behavioral activation (BA) focuses on constructing reinforcement contingencies that increase functional behavior and self‐efficacy. Self‐efficacy refers to the person’s beliefs about the abilities to exercise control on the events affecting their lives. The intervention was adapted from the Chinese Version of the Coping with Caregiving.	General discussion about the psychoeducation program and information packet.
WHAT materials	Written information, including the forms for pleasant event scheduling, was mailed to participants before the program started.	Information packet.
PROCEDURES	Four weekly sessions of psychoeducation and eight sessions of behavioral activation. Themes of the psychoeducation component were as follows. Session 1 (week 1): · Symptoms and associated behavioral changes in dementia · Stages in dementia · Caregiving roles and demands · Effects on caregivers Session 2 (week 2): · Physical, social and psychological consequence of stress · Identifying stress reactions · Awareness of stress · Stress and well‐being Session 3 (week 3): · The effect of life events on mood · Tracking daily/ weekly events · Identifying pleasant events · Scheduling pleasant events Session 4 (week 4): · Communication needs to family members · Types of communications: passive, aggressive and assertive · Resources available in the community · Planning in the future Themes of the behavioral activation component were: · Session 1 – Review the present use of time. Using the monitoring form · Session 2 – Brain‐storm pleasant events. Scheduling pleasant activities · Session 3 – Review scheduling of events. Discuss how to improve · Session 4 – Review modifications. Consolidate gains on scheduling · Session 5 – Review present social support. Explore new sources of support · Session 6 – Examine communication skills. Explore new options · Session 7 – Review new communications. Discuss how to improve · Session 8 – Review modification. Consolidate gains on support	All TGM participants received four weekly psycho‐education sessions over the phone with the same contents as in the TBA group. These caregivers were then assigned to eight bi‐weekly sessions of general monitoring with no BA intervention. Each of these sessions started with checking in with the caregiver through inviting them to update their caregiving situation. Caregivers were then guided to discuss one of the following topics at each session in this order: 1. caregiver’s health 2. care‐recipient’s needs 3. caregiver’s routines 4. social support. As there were a total of eight sessions, the last four sessions repeated the order of the first four. While some caregivers might report on attempts they made on their own initiative to improve their scheduling and communication, no specific attempt was made to ask them to review these attempts.
WHO provided	An interventionist with a degree in social work delivered all the four sessions of psycho‐education. Six paraprofessional coaches, between 50 and 60 years old and with an undergraduate degree in helping or service professions, delivered the BA or MG interventions, or carried out monitoring. A social worker and a clinical psychologist provided the training and facilitated weekly supervision separately for TBA and TGM coaches.
HOW delivered	Individually over the telephone.
WHERE occurred	At home.
WHEN and HOW MUCH	Twelve sessions: sessions 1 to 4 were delivered weekly and sessions 5 to 12 were delivered biweekly. Each session lasted about 20 minutes.
TAILORING	Interventions were comprehensively designed, not tailored to cover individual or unmeet needs.
MODIFICATIONS	None described.
HOW WELL planned	The program fidelity was assessed by a rating system built into recording form. At the end of each session, all interventionists, including the paraprofessionals, were asked to rate to what extent they were able to follow the protocol for each of the four PE sessions (3 = fully; 2 = adequately with at least 60% of the material covered; 1 = slightly; 0 = not at all). A similar procedure was adopted for each of the 8 sessions for both TBA and TGM. In addition, 10 cases from TBA and 10 cases from TGM were audiotaped. Interventionists’ adherence to the intervention protocol was assessed by two graduate students who had received eight hours of training on the coding scheme. The sessions were coded with reference to four core TBA strategies (activity planning, review to improve on scheduling, develop new help‐seeking communication skills and review to improve on communications) and four core TGM strategies (updating on caregiving situation, health and needs of the caregiver and the care‐recipient, daily routines and family communications).
HOW WELL actual	As planned. Minimal deviations to the main components of the intervention.

Table 3. TIDieR ‐ Blom 2015

Study	Blom 2015
TIDieR item	Experimental intervention	Control intervention
BRIEF NAME	Mastery of dementia (MoD).	Minimal Intervention (e‐bulletins).
WHY	A guided self‐help Internet course ‘Mastery over Dementia’ (MoD) designed to reduce caregivers’ symptoms of depression and anxiety. MoD included a combination of psycho‐education with active participation of the caregiver, management of behavioral problems, teaching coping strategies, components of cognitive behavioral therapy (cognitive reframing: changing non‐helpful into helpful thoughts), and increasing social support.	No applicable.
WHAT materials	Internet contents.	E‐bulletins (digital newsletters).
PROCEDURES	The Internet course consists of 8 lessons and a booster session with the guidance of a coach monitoring the progress of participants and evaluating the homework. Each lesson has the same structure and consists of information (text material and videos), exercises, and homework, with an evaluation at the start and end of each session. The elements of the course were presented in the following order: · coping with behavioral problems (problem solving) · relaxation · arranging help from others · changing non‐helping thoughts into helping thoughts (cognitive restructuring) · communication with others (assertiveness training). The booster session was provided a month after participants finished the eight lessons, and provided a summary of what was learned. After every lesson, participants sent their homework to a coach via a secure application. The coach sent electronic feedback to caregivers on their homework within three working days. The feedback had to be opened before the next lesson could be started. Participants were automatically reminded to start with a new lesson or to send in their homework if they were not active for a fixed period of time. All participants in this study received feedback from the same coach, a psychologist employed by a health care agency with additional training in cognitive behavioral therapy and experience in the field of dementia.	Caregivers received a minimal intervention consisting of e‐bulletins (digital newsletters) with practical information on providing care for someone with dementia. The bulletins were sent by email according to a fixed schedule (every 3 weeks) over nearly 6 months. The topics of the bulletin, which did not overlap with the content of MoD, were: · driving · holiday breaks · medication · legal affairs · activities throughout the day · help with daily routines · grieving · safety measures in the home · possibilities for peer support. There was no contact with a coach over the length of the study.
WHO provided	Internet intervention. A coach monitored the progression of the participants.	No applicable.
HOW delivered	Online.
WHERE occurred	At home.
WHEN and HOW MUCH	Eight lessons plus a booster session.	The bulletins were sent by email according to a fixed schedule (every 3 weeks) over nearly 6 months.
TAILORING	Interventions were comprehensively designed, not tailored to cover individual or unmeet needs.
MODIFICATIONS	None described.
HOW WELL planned	No described.
HOW WELL actual	No described.

Table 4. TIDieR ‐ Brennan 1995

Study	Brennan 1995
TIDieR item	Experimental intervention	Control intervention
BRIEF NAME	ComputerLink.	None beyond general information to identify local resources.
WHY	The intervention aimed to help caregivers develop problem‐solving skills, manage their own emotions, and increase their knowledge of dementia and caregiver‐support strategies.	No applicable.
WHAT materials	Caregivers received a Wyse 30 terminal and a Everex modem system.	No applicable.
PROCEDURES	ComputerLink provided three functions: · information · decision support · communication The electronic Encyclopedia provided extensive factual information to enhance self‐care and understanding of Alzheimer’s disease and promote health management of the care recipient. The program uses a system to help caregivers to resolve problems basing in the multiattribute utility theory of von Winterfeldt & Edwards. ComputerLink permitted several options for communication among caregivers, including private mail, access to a public bulletin board, and a section of questions and answers. A nurse moderator entered the system daily and read and responded to messages. The nurse served as a moderator and facilitator.	None beyond general information to identify local resources.
WHO provided	Study nurses who had completed an intervention training programme.	No applicable.
HOW delivered	Online.	No applicable.
WHERE occurred	At home.	No applicable.
WHEN and HOW MUCH	Participants had access to the system 24 hours a day during 12 months.	No applicable.
TAILORING	Intervention was comprehensively designed, not tailored to cover individual or unmeet needs.	No applicable.
MODIFICATIONS	None described.	No applicable.
HOW WELL planned	Indicators for logging in the system.	No applicable.
HOW WELL actual	Caregivers logged into the system 3,875 times, They accessed it a mean of 83 times over 12 months. A typical encounter lasted 13 minutes.	No applicable.

Table 5. TIDieR ‐ Cristancho‐Lacroix 2015

Study	Cristancho‐Lacroix 2015
TIDieR item	Experimental intervention	Control intervention
BRIEF NAME	Diapason.	Treatment as usual.
WHY	The program’s content on the web was based on cognitive theories of stress.	No applicable.
WHAT materials	Web materials.	No applicable.
PROCEDURES	· Session 1: Caregiver stress – this session presents a definition of stress, its causes and consequences on caregivers, risk factors for chronic stress, and mechanisms and effects of relaxation (includes a link to the relaxation training in the Diapason website), as well as strategies for managing stress underlining the importance of looking for respite. · Session 2: Understanding the disease – in this session, the Alzheimer’s disease diagnosis procedure, the symptoms, the progression of the illness, and the consequences on daily life activities for persons with Alzheimer’s disease (PWAD) are explained. · Session 3: Maintaining the loved ones’ autonomy – this session presents the reasons and strategies to involve loved ones in the process of care in order to stimulate the preserved functions and compensate for the lost ones. The session underlines the importance of maintaining the self‐esteem of PWAD. · Session 4: Understanding their reactions – in this session, the most frequent behavioural and psychological symptoms of dementia (BPSD) and their characteristics are succinctly described and illustrated by examples from daily life. The contextual and intrinsic factors that might be associated with them are also described. · Session 5: Coping with behavioral and emotional troubles – this session presents practical advice on how to cope vis‐à‐vis the BPSD described in the previous session. · Session 6: Communicating with loved ones – this session includes the description of the most frequent language troubles and the strategies to modulate and adapt communication to the preserved skills of PWAD. · Session 7: Improving their daily lives – this session presents strategies to facilitate the performance of activities that become difficult or impossible to execute due to apraxia, illustrating them with examples adapted to daily life. · Session 8: Avoiding falls – the session includes practical advice for maintaining and stimulating the relative’s balance and actions to adopt in the event of a fall. In addition, various actions are described to adapt the relative’s home. · Session 9: Pharmacological and non‐pharmacological interventions – this session includes a brief presentation of different interventions available for caregivers in France with pharmacological treatment as well as cognitive and psychological support. · Session 10: Social and financial support – this session presents the different stakeholders and services that may help caregivers in their daily life. The financial and social support provided by the French government is also revised. · Session 11: About the future – this session provides caregivers with information about the role of disease progression anticipation, inviting them to try and foresee solutions keeping a prospective vision, encouraging them to look for further sources of information, and social support to reduce the uncertainty of caregiving situations. · Session 12: In a nutshell – the last session encompasses a summary of the Diapason program, emphasizing the acceptance of support and help and the importance of obtaining more information to anticipate and avoid stressful circumstances. Additionally the website contains other sections that can be consulted at any time: · Relaxation training: guidelines for learning relaxation as well as 2 videos for the modelling of Schultz’s Autogenic Training and Jacobson’s method. · Life Stories: stories about 4 couples, based on testimonials of caregivers, in which difficult situations are illustrated and possible solutions to manage them are discussed (i.e., apathy of patient, caregivers’ isolation). · Glossary: a glossary for technical words (i.e., neuropsychological assessment, aphasia). · Stimulation: practical activities to stimulate autonomy and share pleasant activities with the relatives in daily life. · Forum: a private and anonymous forum to interact with peers, to express their concerns, discuss solutions to daily problems, and share their feelings and experiences. The participants use nicknames to protect their privacy. A clinical psychologist participates in the discussions if necessary (i.e., avoiding aggressive or inappropriate comments).	No applicable.
WHO provided	No web‐moderate self‐provided intervention except when using the forum resource.	No applicable.
HOW delivered	Online.	No applicable.
WHERE occurred	Home.	No applicable.
WHEN and HOW MUCH	The length of the intervention was 3 months, with each weekly session lasting 15 to 30 minutes on average, but there was no time limit and the participants could access different website sections (i.e., relaxation training, forum) for as long as they wished at any time.	No applicable.
TAILORING	Intervention was comprehensively designed, not tailored to cover individual or unmeet needs.	No applicable.
MODIFICATIONS	None described.	No applicable.
HOW WELL planned	Indicators for login in the system.	No applicable.
HOW WELL actual	On average, participants used the website 19.7 times (SD 12.9) and for 262.2 minutes (SD 270.7) during the first 3 months.	No applicable.

Table 6. TIDieR ‐ Czaja 2013

Study	Czaja 2013
TIDieR item	Experimental intervention	Control intervention
BRIEF NAME	Named "intervention condition".	Two control conditions which were combined: attention control and information
WHY	The intervention condition was modelled after the REACH II intervention.	No described.
WHAT materials	A videophone (CISCO IP 7900) was installed in the caregivers’ homes and connected to a Digital Subscriber Line and to a secure server at the host site. Participants in this condition also received a caregiver notebook with basic information about caregiving and community resources.	Caregivers in the attention condition were provided with a notebook with information related to nutrition. The information condition group received educational materials with basic information about dementia, safety, and community resources.
PROCEDURES	The intervention included education and skills training. It was designed to address five caregiver risk areas (safety, social support, problem behaviours, depression, and caregiver health). The intervention was standardized with respect to the treatment domains and strategies used within the treatment domains but individualized with respect to the amount of emphasis within a treatment domain based on a risk assessment administered at baseline. Certified interventionists taught problem solving strategies to deal with the care recipients’ problem behaviours and training on stress management, healthy behavior strategies, community resources, and communication strategies. The intervention included six 1‐hour monthly sessions. The four educational seminars were brief video lectures from experts on topics relevant to caregiving (i.e., update on Alzheimer disease, caregiver depression) and were presented serially; a new video appeared monthly beginning in month 2 of the intervention.The intervention also included five videophone support group sessions, which were interspersed throughout the intervention period. The support groups were structured and included up to six caregivers and a certified group leader. During the support group sessions caregivers received topical information (related to the material covered in the individual intervention sessions) and shared experiences and concerns.	Attention condition: participants received two 1‐hour in‐home sessions and four videophone sessions. They also participated in five telephone support group sessions, which were interspersed with the individual intervention sessions. The support groups followed the same format as those for the intervention group. The content for the attention control condition was structured around nutrition and healthy eating. Information condition: participants were mailed a packet of educational materials and received a brief (<15 minute) telephone “check‐in call” at 3 months post randomisation.
WHO provided	Interventionists. All assessors and interventionists received training and were certified before entering the field.
HOW delivered	Online plus telephone.	Face to face plus telephone.
WHERE occurred	At home.
WHEN and HOW MUCH	6 sessions (1 hour each).	· 6 sessions (1 hour each). · 15 minutes “check‐in‐call”.
TAILORING	Interventions were comprehensively designed, not tailored to cover individual or unmeet needs.
MODIFICATIONS	None described.
HOW WELL planned	Interventionists submitted taped treatment sessions for review and feedback as part of the certification. Treatment implementation was monitored and maintained by weekly supervision meetings. Both interventionists and assessors followed a detailed manual of operations and a delivery assessment form was completed after each contact.
HOW WELL actual	No described.

Table 7. TIDieR ‐ Davis 2994

Study	Davis 2004
TIDieR item	Experimental intervention	Control intervention
BRIEF NAME	Telephone training or in‐home training (this review only includes evidence from telephone training).	Friendly, socially supportive, telephone calls.
WHY	The study was grounded in the stress, appraisal and coping theory of Lazarus. The premise for the caregivers skill training intervention was that improvements in caregiver's outcomes could be addressed by expanding caregiver's repertoire of effective appraisal and coping strategies through telephone training strategies	No applicable.
WHAT materials	Caregivers received a standardized, loose‐leaf notebook with suggestions for managing frequently encountered Alzheimer’s disease situations. They had to fill out a weekly log on problems they had managed over the past week.	Socially supportive friendly telephone calls.
PROCEDURES	Caregivers participated in training sessions focused on: · general problem solving · caragiver appraisal of behavior problems · written behavioral programs for managing specific problems · strategies for handling effective responses to difficult caregiving situations. They received an initial home visit of 45 minutes to introduce the trained interventionist and to familiarize with the uses of the notebook and log sheets. At each weekly contact, the interventionist reviewed the past week of caregiving problems with the caregiver, discussed the log entries and used selected notebook sections to help the caregiver manage the situation.	Friendly callers made an initial home visit of approximately 45 minutes to introduce themselves. At the time of each contact, callers required about the caregiver's past week as well as any changes in their general health and medication regimen.
WHO provided	Interventionists.
HOW delivered	Telephone.
WHERE occurred	At home.
WHEN and HOW MUCH	12 weekly telephone contacts.
TAILORING	Interventions were comprehensively designed, not tailored to cover individual or unmeet needs.
MODIFICATIONS	No described.
HOW WELL planned	Telephone contacts were tape recorded and reviewed to assess continued adherence.
HOW WELL actual	Telephone contacts averaged 37 minutes (SD=18)	Friendly telephone calls averaged 16 minutes (SD=12).

Table 8. TIDieR ‐ Dowling 2014

Study	Dowling 2014
TIDieR item	Experimental intervention	Control intervention
BRIEF NAME	LEAF (Life Enhancing Activities for Family Caregivers).	Attention control group.
WHY	The positive affect intervention teaches participants a series of behavioral and cognitive “skills” for increasing positive affect coping with health‐related and other life stresses. The skills include noticing and capitalizing on positive events, gratitude, mindfulness, positive reappraisal, focusing on personal strengths, attainable goals, and acts of kindness.	No described.
WHAT materials	No described.
PROCEDURES	Skill‐building Intervention Sessions · Session 1: Positive Events and Gratitude – Identify a positive or meaningful event in the last week and what it means to note, favour, and capitalize on positive events. Homework: write down 3 things that went well each day and why they went well. Begin a gratitude journal (writing one thing each day you are grateful for) that continued throughout the rest of the intervention. · Session 2: Mindfulness – Teach concepts of mindful attention and non‐judgment. Homework: (a) practice awareness of breathing and meditation for 10 minutes daily and (b) once a day take the time to enjoy something that you usually hurry through, do one thing at a time, and pay attention to it. Participants were encouraged to continue the breathing exercise through the remaining weeks of the intervention and to continue the gratitude journal. · Session 3: Positive Reappraisal – Discuss the meaning of positive reappraisal and how to apply it to everyday occurrences. Homework: each day think of one negative or stressful thing that happened. Practice positive reappraisal – why it may not be as bad as initially thought or something good that might come of it. Participants were encouraged to write about their experience reappraising at the end of each day and to continue their gratitude journal and mindful breathing exercises. · Session 4: Personal Strengths and Attainable Goal – Generate a list of personal strengths that can be used in everyday life. Define attainable goals and practice setting one related to self‐care. Homework: achieve attainable goal. Participants were encouraged to write about their goals at end of day and to continue their gratitude journal and mindful breathing exercises. · Session 5: Altruistic Behaviors/Acts of Kindness – Doing for Others – Discuss the positive impact of doing for others. Homework: do something nice for someone else each day. Participants were encouraged to write about their acts of kindness at end of day and to continue their gratitude journal and mindful breathing exercises.	Participants randomised to the control group engaged in 5 one‐on‐one sessions with a facilitator. The sessions were comparable in length to the intervention sessions (approximately 1 hour) but consisted of an interview and did not have any didactic portion or skills practice. Each session began with the completion of the modified Differential Emotions Scale (DES). In addition to these affect questions, each session had qualitative and quantitative questions and activities cantered on a theme (i.e., life history, health history, diet and exercise, social networks, and meaning and spirituality), to keep the sessions different and interesting for participants. Home practice for the control group consisted of the brief daily affect reports. At the start of sessions 2 through 6, the facilitator reviewed the previous week’s affect diary with the participant.
WHO provided	Facilitators for the intervention sessions were trained for content and delivery. Two were clinical nurse specialists in fronto‐temporal dementia and 1 was a psychologist.	The facilitator for the control group sessions was a research associate, familiar with the content of the intervention sessions so as to not engage in the intervention content during control sessions with participants.
HOW delivered	Mainly by video‐conference (only 1 subject participated in‐person, all others participated remotely).
WHERE occurred	At home.
WHEN and HOW MUCH	6 sessions of 1 hour each.
TAILORING	Interventions were comprehensively designed, not tailored to cover individual or unmeet needs.
MODIFICATIONS	None described.
HOW WELL planned	All sessions were audio‐recorded digitally for both quality assurance and intervention evaluation.
HOW WELL actual	No described.

Table 9. TIDieR ‐ Duggleby 2018

Study	Duggleby 2018
TIDieR item	Experimental intervention	Control intervention
BRIEF NAME	My Tools 4 Care (MT4C).	Information.
WHY	The aim of the intervention, based on transition theory, was to support carers through transitions and increase their self‐efficacy, hope, and health related quality of life. The intervention had multiple components and included choice, as the carers could choose which sections they would like use, and when. User‐generated content is encouraged throughout the intervention as participants may write in sections, add stories, pictures, music etc. Sections include: · About Me · Common Changes to Expect · Frequently Asked Questions and Resources · Important Health Information (about person they are caring for).	No applicable.
WHAT materials	An online version of the Toolkit entitled My Tools 4 Care (MT4C) (https://www.mytools4care.ca/).	A copy of the Alzheimer Society’s booklet The Progression of Alzheimer’s Disease, sent by email.
PROCEDURES	Participants received information for the MT4C website, (i.e., the website address, a unique username and password to access the site for 3 months), and an electronic copy of a Toolkit Checklist in which the participant will be asked to document their use of MT4C (i.e., time spent and content accessed over the 3 months). Participants also received an electronic copy of the Alzheimer Society’s educational booklet, The Progression of Alzheimer’s Disease.	The electronic document consists of a five‐part series that describes the stages of Alzheimer’s disease. It is written for the person with Alzheimer disease, their family and carers and is freely accessible via the Alzheimer Society of Canada website.
WHO provided	Whole intervention was over the Internet.	No applicable.
HOW delivered	Online	Emailed booklet.
WHERE occurred	At home.
WHEN and HOW MUCH	Intervention was accessible for 3 months.	No applicable.
TAILORING	Interventions were comprehensively designed, not tailored to cover individual or unmeet needs.
MODIFICATIONS	None described.	No applicable.
HOW WELL planned	Webpage recorded the time and duration of access.	No applicable.
HOW WELL actual	73% of caregivers used MT4C at least once over the 3‐month period. By 3 months, participants spent most of their time on Section 2 – Common changes to expect (median 15 minutes) and Section 4 – Resources (median 10 minutes).	No applicable.

Table 10. TIDieR ‐ Gant 2007

Study	Gant 2007
TIDieR item	Experimental intervention	Control intervention
BRIEF NAME	Video/workbook/telephone coaching condition.	Education/check‐in‐call condition.
WHY	No described.
WHAT materials	Caregivers in the video condition received a 10‐session video series, a workbook from the Dementia Caregiving Skills Program, and weekly telephone calls from a trained coach.	Participants received by mail a 37‐page booklet, Basic Dementia Care Guide and phone calls.
PROCEDURES	The intervention used primarily behavioural strategies: behavioural activation, behavioural management, and stress reduction through relaxation training. The workbook provided didactic and experimental materials to reinforce information presented in each video session. Participants received 12 weekly phone calls by trained research staff who served as coaches. The first 10 calls reinforced each of the video sessions, the last 2 calls served as follow‐up for further application of concepts. Coaches followed a coach manual that provided a script for reviewing didactic materials and assignments with caregivers, and for assisting them in the application of intervention concepts to their problems. During the coaching calls caregivers reported the specific behavioural strategies that they devised, written down, used, and evaluated, based on the behaviour management module that they learned to apply to their situations.	Intervention included information on dementia and suggestions for dealing with a variety of caregiving challenges. In a cover letter, procedures for maximizing the benefits of this educational booklet were provided. Caregivers then received approximately 7 biweekly telephone calls by a trained staff member. In these calls, the staff member checked on the safety of the caregiver and family member, discussed the caregiver’s use of the suggestions from the guide, and responded to questions by referring the caregiver to appropriate sections in the guide. A standardized script was used for calls to participants in this comparison condition.
WHO provided	Coaches following a coach manual.
HOW delivered	By telephone.
WHERE occurred	At home.
WHEN and HOW MUCH	12 weekly telephone calls.	7 bi‐weekly telephone calls.
TAILORING	Interventions were comprehensively designed, not tailored to cover individual or unmeet needs.
MODIFICATIONS	None described.
HOW WELL planned	No described.
HOW WELL actual	No described.

Table 11. TIDieR ‐ Gustafson 2019

Study	Gustafson 2019
TIDieR item	Experimental intervention	Control intervention
BRIEF NAME	D‐CHESS (Dementia–Comprehensive Health Enhancement Support System).	Caregiving booklet.
WHY	D‐CHESS was designed to help with motivation, decision making, stress reduction, and access to services by allowing caregivers to obtain information and support without concern for location, distance, time, confidentiality, or education.	No described.
WHAT materials	Website. Participants were also offered the use of three commercially available sensors: a Bluetooth tracker, a GPS location tracker, and a motion sensor. The Bluetooth tracker would help locate lost items such as keys; the GPS location tracker could be carried by a patient, allowing the caregiver to monitor the patient’s location; and the motion sensor could be used to alert a caregiver if the patient entered specific areas of the home that could be hazardous.	No applicable.
PROCEDURES	This system offered detailed information about dementia, personal accounts of how other caregivers coped, descriptions and locations of services that caregivers may need, and links to vetted online resources. Users could anonymously pose questions to experts, receive suggestions of resources for help, communicate with other caregivers through facilitated discussion groups, and use decision aids to work through key issues. In addition, D‐CHESS services offer other elements proven effective in caregiver programs, such as assertive out‐reach, action planning, and peer and family support, as well as quality informational resources such as expert advice and vetted articles and websites.	No described.
WHO provided	Whole intervention was over the Internet.	No described.
HOW delivered	Online.	Booklet.
WHERE occurred	At home.
WHEN and HOW MUCH	The intervention group received access to the D‐CHESS website for 6 months.	No applicable.
TAILORING	Interventions were comprehensively designed, not tailored to cover individual or unmeet needs.
MODIFICATIONS	None described.
HOW WELL planned	D‐CHESS automatically collected use data, including: number of participants logging on, number of page views, specific services used, discussion group posts, data input to decision aids, and private messages sent to the Alzheimer’s disease information specialist at the Wisconsin Alzheimer’s Disease Research Center.	No applicable.
HOW WELL actual	Overall, participants continued to access D‐CHESS throughout the study, with 100%, 50%, 71%, 71%, 57%, and 64% logging on during months 1 through 6, respectively.	No applicable.

Table 12. TIDieR ‐ Hattink 2015

Study	Hattink 2015
TIDieR item	Experimental intervention	Control intervention
BRIEF NAME	STAR E‐Learning (European Skills Training and Reskilling).	Waiting list.
WHY	The STAR training portal was designed to provide dementia care training both for informal and formal caregivers. It is designed as a multilingual e‐learning tool from experts on education, technology, and dementia care.	Participants in the control group were informed they could follow the e‐learning course after post‐test measurements were performed within the experimental intervention.
WHAT materials	The course content was developed from 3 perspectives: · Medical model of dementia: information on types of dementia, symptoms and diagnostics · Functional consequences in daily life and how to compensate for disabilities · Dealing with the psychosocial consequences for the person with dementia and his family	No applicable.
PROCEDURES	The STAR training portal offers the following functionality: · 8 modules on different topics in dementia care (2 at basic level, 6 at intermediate to advanced level): what is dementia; living with dementia; getting a diagnosis; practical difficulties in daily life; emotional impact of dementia; support strategies; positive and empathic communication; looking after yourself · Learning Path Advisor to assess baseline knowledge and confidence, to suggest the point to start the course · Facebook and LinkedIn communities to promote peer support and provide opportunities to contact dementia care professionals The modules consist of text, videos, interactive exercises, knowledge tests, and also include references to other websites, literature, and videos.	No applicable..
WHO provided	Whole intervention was over the Internet.	No applicable.
HOW delivered	Online.	No applicable.
WHERE occurred	At home.
WHEN and HOW MUCH	The intervention group received access to the STAR website for 4 months.	No applicable.
TAILORING	Interventions were comprehensively designed, not tailored to cover individual or unmeet needs. Caregivers can follow their own learning path.
MODIFICATIONS	None described.
HOW WELL planned	Pilot study to assess STAR’s usefulness, user friendliness, and impact on knowledge.	No applicable.
HOW WELL actual	As planned.	No applicable.

Table 13. TIDieR ‐ Hayden 2012

Study	Hayden 2012
TIDieR item	Experimental intervention		Control intervention
BRIEF NAME	The use of Internet technology for psychoeducation and support with dementia caregivers.		Customary care plus monthly brief telephone calls.
WHY	Data suggest that providing education, social support, and ongoing professional consultation to families involved in the care of a relative with Alzheimer’s disease results in improvement in caregiver psychological status, and sometimes even slows the functional decline of the patient.		No applicable.
WHAT materials	Customary care and access to an intensive, interactive online education and support website.		Customary care and telephone calls.
PROCEDURES	· Online education and support, provided by nine brief educational streaming videos, with brief written brochures, plus non‐compulsory weekly 1‐hour professionally hosted synchronous online chats to review the educational material and problem‐solving care issues, using the rest of the group as sources of feedback. Participants received weekly telephone and email reminders about the chat sessions. · Additional focus cantered on self‐care promotion via behavioural interventions and goal attainment scaling.	Active control condition involving supportive phone calls.
WHO provided	Geriatricians, a geriatric psychiatrist, clinical psychologists, a research sociologist, and master’s and bachelor’s level support staff. Both interventions were provided by the same team.
HOW delivered	Internet.		Telephone.
WHERE occurred	At home.
WHEN and HOW MUCH	6‐month interventions.
TAILORING	Interventions were comprehensively designed, not tailored to cover individual or unmeet needs.
MODIFICATIONS	None described.
HOW WELL planned	No described.
HOW WELL actual	Reports of technological challenges encountered by many of the participants as well as constraints due to VA resources and information security policies.

Table 14. TIDieR ‐ Huis in het Veld 2020

Study	Huis in het Veld 2020
TIDieR item	Experimental intervention	Control intervention
BRIEF NAME	Major Self‐Management Support Intervention.	· Medium Self‐Management Support Intervention · Minor Self‐Management Support Intervention
WHY	The intervention is based on the 5 steps of the “5A model” of self‐management support and the person‐centered care theory of Kitwood. · assessing the state of behavior, beliefs, and motivation · advising based upon personal health risks · agreeing on a realistic set of goals · assisting in anticipating barriers and developing a specific action plan · arranging follow‐up.	Not described.
WHAT materials	Emails, online videos, e‐bulletins.	Online videos and e‐bulletins.
PROCEDURES	The major intervention arm consists of the following elements: · 3 personal email contacts with a nurse specialized in dementia care · provision of 6 online videos about how to manage behavior changes in a relative with dementia and to improve self‐efficacy in managing with this behavior · provision of e‐bulletins with practical information about different types of behavior changes and how to manage them The nurse gave feedback on assignments and gave feedback on the plan that the family caregiver came up with in the assignments.There were 6 videos about different types of behavior changes that occur frequently (dependent behavior, aggressive behavior, suspicious behavior, apathy or indifference, nighttime restlessness, and masking behavior). Family caregivers chosen the number of videos they watch and the accompanying assignments that they do themselves, depending on their own needs and the behavior changes that occur in their relative with dementia. The behavior changes covered in the bulletins were the same as in the videos. The e‐bulletins involve assignments to help caregivers translate the generic information to their own situation and to reflect on possible causes of the behavior changes, how they want to influence the behavior, and how they want to cope with it.	· The medium self‐management support intervention consists only of the online videos and e‐bulletins · The minor self‐management support intervention consists only of the e‐bulletins, the same as those in the major and medium support interventions
WHO provided	The personal email contacts were handled by a nurse with a Bachelor’s or Master’s qualification in nursing and with follow‐up training in dementia care.
HOW delivered	Email plus online videos plus e‐bulletins.	Online videos plus e‐bulletins.
WHERE occurred	At home.
WHEN and HOW MUCH	6 weeks.
TAILORING	Interventions were comprehensively designed, not tailored to cover individual or unmeet needs.
MODIFICATIONS	None described.
HOW WELL planned	No described.
HOW WELL actual	No described.

Table 15. TIDieR ‐ Kajiyama 2013

Study	Kajiyama 2013
TIDieR item	Experimental intervention	Control intervention
BRIEF NAME	iCare Stress Management e‐Training Program (ICC).	Education/Information‐Only Condition (EOC).
WHY	The iCare program is an adaptation of the psycho‐ educational program ‘Coping with Caregiving’ (CwC) developed by Gallagher‐Thompson and colleagues that aims to teach a set of core coping skills to caregivers for stress management. These include: relaxation training; learning to increase everyday pleasant activities (including if possible some pleasant activities with the person with dementia); cognitive restructuring (of unduly negative appraisals regarding caregiving and themselves); and improved communication skills on how to ask for help effectively from family members and relevant community and medical institutions.	The EOC served as a control for the purpose of obtaining new information about dementia and what to expect in caring for dementia patients. It also controlled for the involvement with relevant information on the Internet and for the general cognitive and experiential capacity required to ‘navigate’ in complex Internet materials.
WHAT materials	Website and video‐recordings.
PROCEDURES	Educational modules had a hierarchical structure. Participants were encouraged to practice specific assignments in each module over a 7‐to‐10‐day interval before moving to the next one. The iCare program consists of the following modules: · information segment about what ‘dementia’ means and what are common problems associated with it · components on dealing with stress including techniques for relaxation, stress management, and challenging unduly negative thoughts about caregiving · behavioral activation (increasing everyday positive activities for oneself and the patient with dementia) · communication skills to improve help‐seeking with family and community institutions as well as improving ability to relate to the patient with dementia · managing difficult behaviours of the patient with dementia · review of ‘healthy habits’ (nutrition and exercise) for the caregiver along with information on national resources they can consult for further on‐going assistance At the end of each module, CGs were asked to create their own individual ‘action plans’ in the workbook requiring them to describe what they had learned by completing the module and relevant assignments and how they planned to use the information from that specific module in the future in their daily lives.	Caregivers assigned to the EOC were exposed to a website containing similar navigational features, but with content focused on information about dementia, obtained from reputable national sources. In addition, links to certain video‐taped information were provided.
WHO provided	Online.
HOW delivered	Online.
WHERE occurred	At home.
WHEN and HOW MUCH	During 3 months.	No applicable.
TAILORING	Interventions were comprehensively designed, not tailored to cover individual or unmeet needs.
MODIFICATIONS	None described.
HOW WELL planned	No described.
HOW WELL actual	No described.

Table 16. TIDieR ‐ Kwok 2013

Study	Kwok 2013
TIDieR item	Experimental intervention	Control intervention
BRIEF NAME	Psychoeducational intervention.	DVD information.
WHY	The intervention is based on theories of psychosocial transition and stress coping, under the framework of cognitive behavioral therapy, and focuses on: · providing emotional support · directing caregivers to appropriate resources · encouraging caregivers to attend to their own physical, emotional, and social needs · educating caregivers on strategies to cope with ongoing problems	No described.
WHAT materials	Manualized psychoeducational intervention.	DVD with educational information about dementia caregiving.
PROCEDURES	The participants in the intervention group were educated and given advice on topics related to dementia caregiving: · knowledge of dementia · skills of communicating with the patient · management of behavioral and psychological symptoms of dementia · caregivers’ own emotional issues · resources available in the community · long‐term care plan The topics covered and the schedule of presentation were similar to typical psychoeducation interventions held “on site” at community centres.	No applicable.
WHO provided	A registered social worker.	No applicable.
HOW delivered	By telephone.	DVD.
WHERE occurred	At home.
WHEN and HOW MUCH	12 sessions (approximately 30 minutes per session, one session per week).	No applicable.
TAILORING	Interventions were comprehensively designed, not tailored to cover individual or unmeet needs.
MODIFICATIONS	None described.
HOW WELL planned	No described.
HOW WELL actual	No described.

Table 17. TIDieR ‐ Mahoney 2003

Study	Mahoney 2003
TIDieR item	Experimental intervention		Control intervention
BRIEF NAME	TLC System.		Control.
WHY	The theoretical model asserts that the background and context of stress (caregiving relationship), plus the stressor (care recipient’s behaviours), as mediated by caregiver mastery, resulted in manifestations of caregiver stress.		No described.
WHAT materials	The technology consisted of an integrated telephone network system and an IVR computer network system.		No applicable.
PROCEDURES	Caregivers dialled in from any standard touch‐tone telephone and heard the narrator greet them by name, review the menu of four module options, and provide the service they requested. The included modules were: · Weekly caregiver’s conversation: monitored the caregiver’s stress levels and provided information on how to manage the care recipient’s behavioural problems. · Personal mailbox: allowed caregivers to send and receive communications through voice mail among themselves or to communicate with a clinical nurse specialist who directly answered or triaged questions to a multidisciplinary professional panel of Alzheimer’s disease experts. · Bulletin board: in‐home telephone support group that provided personal and caregiver group voice mail, similar to a computer chat group. · Activity‐respite conversation: care recipient distraction module designed to reduce disruptive behaviours and to provide caregivers with respite time.		No applicable.
WHO provided	Internet intervention.		No applicable.
HOW delivered	Telephone‐delivered.		No applicable.
WHERE occurred	At home.
WHEN and HOW MUCH	The intervention group had access to the system for a 12‐month period, after which their passwords were disabled.		No applicable.
TAILORING	Intervention was comprehensively designed and not tailored to cover individual or unmeet needs. However, because caregivers have diverse needs and preferences, the intervention offered multiple components with flexibility to effectively appeal to a variety of users. Participants chose the type of component, frequency, duration, and timing of the usage.	No applicable.
MODIFICATIONS	None described.
HOW WELL planned	No described.
HOW WELL actual	No described.

Table 18. TIDieR ‐ Martindale‐Adams 2013

Study	Martindale‐Adams 2013
TIDieR item	Experimental intervention	Control intervention
BRIEF NAME	Telephone support (CONNECT).	Printed materials.
WHY	Content and structure of the intervention were based on the 6‐month REACH II intervention of 12 individual in‐home and telephone sessions and five telephone support group sessions.	No described.
WHAT materials	Session materials consisted of a Caregiver Notebook and commercially available pamphlets. The Notebook comprised 29 behaviour – management chapters of five to eight pages each (i.e., bathing, repeated questions), and 17 caregiver stress/coping chapters (i.e., assertiveness, communication, grief) based on research and practice, written in large print and at a fifth‐grade reading level.	Control caregivers received pamphlets on dementia and safety as well as telephone numbers for local resources.
PROCEDURES	Like REACH, the multi‐component intervention targeted caregiving risks, including risks associated with emotional and physical well‐being, safety, burden, social support, and patient behaviour management. The first six sessions covered all Caregiver Notebook chapters that were standardized to include the following topics: · introduction · basic information about dementia and financial and legal issues · safety · caregiver health and well‐being · communication · problem solving The sessions were semi‐structured telephone calls with education, skills‐building, and support. Each session began with a relaxation exercise, caregiver updates, review of strategies tried from the preceding session’s topic, and the group leader’s presentation on a behaviour management or stress and coping topic.The rest of the session included discussion and practice by the entire group on the session’s topic, selection of individual strategies to try, selection of the next session’s topic, and closure, including another signal breath relaxation exercise.	Control caregivers received pamphlets on dementia and safety as well as telephone numbers for local resources.
WHO provided	Trained professionals.
HOW delivered	Telephone.	Printed materials.
WHERE occurred	At home.
WHEN and HOW MUCH	Groups met 14 times over 1 year. The support groups met bi‐weekly for 2 months and monthly thereafter for 1 year, for a total of 14 hour‐long sessions.	No described.
TAILORING	Interventions were comprehensively designed, not tailored to cover individual or unmeet needs.
MODIFICATIONS	None described.
HOW WELL planned	No described.
HOW WELL actual	Of the 77 intervention caregivers, 47 (61%) had at least 75% of the 14 sessions and 59 (76.6%) completed at least half. Six caregivers (7.8%) had fewer than three sessions.	No applicable.

Table 19. TIDieR ‐ Metcalfe 2019

Study	Metcalfe 2019
TIDieR item	Experimental intervention		Control intervention
BRIEF NAME	RHAPSODY.		Wait‐list.
WHY	The intervention adopts a multiple‐component model of dementia carer support inspired by the REACH II Study adapted to the care of people with young onset dementia.		No applicable.
WHAT materials	Online device		No applicable.
PROCEDURES	The programme features seven modules: · What is young onset dementia? · Medical perspective on young onset dementia · Frequent problems and solutions · Dealing with challenging behaviours · Family issues · How to get help · Looking after yourself		No applicable.
WHO provided	Internet intervention.		No applicable.
HOW delivered	Online.		No applicable.
WHERE occurred	At home.
WHEN and HOW MUCH	6 weeks.		No applicable.
TAILORING	Intervention was comprehensively designed, not tailored to cover individual or unmeet needs.	No applicable.
MODIFICATIONS	None described.		No applicable.
HOW WELL planned	No described.
HOW WELL actual	No described.

Table 20. TIDieR ‐ NCT00056316

Study	NCT00056316
TIDieR item	Experimental intervention	Control intervention
BRIEF NAME	Behavioural intervention.	Basic education.
WHY	No described.
WHAT materials	Video, workbook and telephone calls.	37‐page Basic Care Guide and telephone calls.
PROCEDURES	Multicomponent behavioral intervention using 10‐session video series, workbook, and weekly telephone coaching sessions.	Participants receive 37‐page Basic Care Guide and bi‐weekly telephone calls by a trained staff member.
WHO provided	Trained staff member.
HOW delivered	Video plus telephone.	Booklet.
WHERE occurred	No described.
WHEN and HOW MUCH	No described.
TAILORING	No described.
MODIFICATIONS	None described.
HOW WELL planned	No described.
HOW WELL actual	No described.

Table 21. TIDieR ‐ NCT03417219

Study	NCT03417219
TIDieR item	Experimental intervention	Control intervention
BRIEF NAME	Education and Skill Building Rehabilitation‐mobile (ESBR‐m).	Usual care plus supplemental educational materials.
WHY	No described.
WHAT materials	Internet contents.	Educational materials.
PROCEDURES	Participants randomised to the ESBR‐m group will participate in four, 90‐minute group ( 5 participants) sessions. These four sessions are supplemented with a booster session one month following the last intervention session.	Participants randomised to the usual care group will receive supplemental educational materials.
WHO provided	No described.
HOW delivered	Internet.	Booklets.
WHERE occurred	At home.
WHEN and HOW MUCH	Four, 90‐minute group sessions.	No applicable.
TAILORING	No described.
MODIFICATIONS	None described.
HOW WELL planned	No described.
HOW WELL actual	No described.

Table 22. TIDieR ‐ Nunez‐Naveira 2016

Study	Nunez‐Naveira 2016
TIDieR item	Experimental intervention		Control intervention
BRIEF NAME	UnderstAID.		Participants in the control group did not use the application and maintained their usual lifestyle.
WHY	No described.		No applicable.
WHAT materials	Smartphone, PC or Tablet.		No applicable.
PROCEDURES	The understAID consists on a Learning section with a database of contents organized in 5 modules with information about 15 different topics. The topics cover information about the care of a person with dementia and caring for oneself as a caregiver, and consist of text, videos, and images and they also include references to other websites. The modules and topics included in understAID are: · Module 1, Cognitive Declines (Topics: Attention, Memory, and Orientation) · Module 2, Daily Tasks (Topics: Bathing, Incontinence, Massage and Touch, and Physical Exercises) · Module 3, Behavioural Changes (Topics: Anxiety and Agitated Behaviour, Depressive Mood, Manic Symptoms, and Emotional Control and Recognition) · Module 4, Social Activities (Communication and Apathy and Loss of Motivation) · Module 5, You as a Caregiver (Topics: Coping with Own Stress and Motivation). Additionally, it has a Social Network section where the caregivers can interact with other participants and exchange information and opinions. This section was moderated by the researchers of the study.		No applicable.
WHO provided	Internet intervention.		No applicable.
HOW delivered	Online.		No applicable.
WHERE occurred	At home.
WHEN and HOW MUCH	Three months.		No applicable.
TAILORING	Intervention was comprehensively designed, not tailored to cover individual or unmeet needs. However, understAID also included the option of filling in an interactive customisation questionnaire with questions about the level of dementia of the person cared by the informal caregiver and about the preferences, energy, and time availability for learning of the caregiver. By answering these questions at the entry of the application, the information level showed to the informal caregivers is personalized and adjusted to their personal situation.	No applicable.
MODIFICATIONS	None described.		No applicable.
HOW WELL planned	No described.		No applicable.
HOW WELL actual	No described.		No applicable.

Table 23. TIDieR ‐ Torkamani 2014

Study	Torkamani 2014
TIDieR item	Experimental intervention		Control intervention
BRIEF NAME	ALADDIN.		The participants in the control group were only assessed at the study three time points, without any further contact or intervention.
WHY	No described.		No applicable.
WHAT materials	Internet contents.		No applicable.
PROCEDURES	ALADDIN is a computerized platform designed to offer avenues of support and information to the carer. It also manages and communicates information related to the patient with dementia and their carers from their home to the clinicians, facilitating distant monitoring. ALADDIN has four key features: · ‘ALADDIN TV’ provides information and educational material about dementia, as well as musical entertainment and relaxation and exercise techniques · The ‘SOCIAL NETWORKING’ feature provides a forum for carers using ALADDIN to communicate with each other · ‘MY TASKS’ is the distant monitoring feature of ALADDIN, where carers complete questionnaires about their own and their relatives’ health. Their responses can subsequently generate clinical alerts based on set parameters, comparing new responses to previous data, resulting in the immediate detection of change by the clinicians monitoring the system · The ‘CONTACT US’ feature allows the carer to alert the clinical site and/or generate a request for contact		No applicable.
WHO provided	Internet intervention.		No applicable.
HOW delivered	Online.		No applicable.
WHERE occurred	At home.
WHEN and HOW MUCH	The project ran for 6 months.
TAILORING	Intervention was comprehensively designed, not tailored to cover individual or unmeet needs. However, being the primary users of ALADDIN, the caregivers chose the schedule of their tasks.	No applicable.
MODIFICATIONS	None described.		No applicable.
HOW WELL planned	The system was monitored twice daily by the clinical teams.		No applicable.
HOW WELL actual	No described.		No applicable.

Table 24. TIDieR ‐ Tremont 2008

Study	Tremont 2008
TIDieR item	Experimental intervention		Control intervention
BRIEF NAME	Family Intervention: Telephone Tracking – Dementia (FITT‐D).		Standard care.
WHY	The intervention is delivered entirely over the telephone. Theoretical underpinnings of the intervention are based on psychosocial transition, transactional stress and coping, and a systems view of family functioning.		No applicable.
WHAT materials	Treatment manual and interventions guide. Telephone.		No applicable
PROCEDURES	Telephone calls included an initial call (orientation and psychoeducation), weekly calls for six weeks, 12 additional contacts every two weeks, and four monthly termination calls. Initial contacts lasted approximately 60 minutes, and follow‐up contacts lasted about 15–30 minutes. The calls focused on providing emotional support, directing caregivers to appropriate resources, encouraging caregivers to attend to their own physical, emotional and social needs, and teaching caregivers strategies to cope with ongoing problems. The intervention did not provide case management, serve as a question‐and‐answer hotline, or provide psychotherapy over the telephone. The FITT method consists of two stages: · The initial stage, Orientation and Psychoeducation, involved providing caregivers with a rationale for the FITT, an introduction to educational and resource materials, a description of what would happen during future phone contacts and an assessment of key areas thought to be instrumental in addressing caregiver burden and mental health (i.e., caregivers’ health, functioning, mood, thinking and family life). The psychoeducation component of this initial stage involved reviewing information about dementia and common psychological, emotional, psychosocial and medical effects of caregiving · The second stage, Follow‐up, involved weekly and bi‐weekly contacts in which new problems were identified, positive and negative changes in caregivers or care recipients were discussed, and psychoeducational information was reviewed and applied for particular situations. The final four follow‐up calls (monthly) addressed issues of termination by allowing caregivers to anticipate FITT contacts coming to an end and to foster reliance on the support network established during the intervention.		No applicable.
WHO provided	Two master’s level therapists who were trained in the FITT‐D procedure delivered the intervention. Doctoral‐level staff supervised therapists weekly to ensure adherence to the protocol and minimize drift.		No applicable.
HOW delivered	Telephone.		No applicable.
WHERE occurred	At home.
WHEN and HOW MUCH	23 telephone contacts over 12 months. Initial contacts lasted approximately 60 minutes, and follow‐up contacts lasted about 15–30 minutes. The intervention involved approximately 12 hours of contact between the therapist and caregiver.		No applicable.
TAILORING	Intervention was comprehensively designed, not tailored to cover individual or unmeet needs.	No applicable.
MODIFICATIONS	None described.		No applicable.
HOW WELL planned	Sessions were audio‐ taped, and two raters reviewed 30 randomly selected telephone contacts and independently completed adherence and competence scales.		No applicable.
HOW WELL actual	No described.		No applicable.

Table 25. TIDieR ‐ Tremont 2015

Study	Tremont 2015
TIDieR item	Experimental intervention	Control intervention
BRIEF NAME	Family Intervention: Telephone Tracking‐Caregiver (FITT‐C).	Telephone Support (TS).
WHY	The theoretical framework of FITT‐C is based on psychosocial transition, transactional stress and coping, and a systems view of family functioning. The three underlying theories of FITT‐C are geared toward enhancing coping within the caregiver through active problem solving and facilitating positive changes within the family system.	The control condition was designed to account for nonspecific therapeutic factors, such as interpersonal contact and relationship. The approach was based on a nondirective control condition.
WHAT materials	Telephone.	Telephone.
PROCEDURES	Focused on providing dementia education, emotional support, directing caregivers to appropriate resources, encouraging caregivers to attend to their physical, emotional, and social needs, and teaching caregivers strategies to cope with ongoing problems. The FITT method consisted of an initial orientation and psycho‐ education call that involved providing caregivers with a rationale of FITT, an introduction to the resource materials, a description of future telephone contacts, and education about dementia and effects of caregiving. The remaining contacts identified changes since the last call, assessed key areas for the caregiver (i.e., health, functioning, mood, social support, and family life), and provided interventions and psychoeducation to help caregivers solve problems and use family resources. The focus of the final two calls changed to identifying helpful aspects of contacts and how these functions could be met after termination. After the final call, the therapist prepared a letter briefly highlighting the progress during the intervention and encouraged the caregiver to continue to develop and use adaptive coping strategies.	The primary goal of this condition was to provide nondirective support for caregivers through empathic and reflective listening and open‐ended questioning. The role of the therapist was to provide unconditional positive regard to caregivers and to establish a supportive relationship. Therapists were discouraged from providing directive strategies, such as education, problem‐solving, advice‐giving, or task directives. However, education was not withheld if the caregiver had the wrong information.
WHO provided	Individuals recruited to serve as therapists had experience working with dementia patients and/or care‐ givers or psychotherapy experience working with adults. Therapists were required to be master’s level and received training in dementia and caregiving. FITT‐C therapists received additional training in the intervention method.
HOW delivered	Telephone.
WHERE occurred	At home.
WHEN and HOW MUCH	16 telephone contacts over 6 months. Although initial telephone contacts had standardized durations (approximately 60 minutes), follow‐up contacts varied depending on the severity of caregiver problems (15–30 minutes).
TAILORING	Interventions were comprehensively designed, not tailored to cover individual or unmeet needs.
MODIFICATIONS	None described.
HOW WELL planned	Quality control was implemented by weekly supervision of both the FITT‐C and TS therapists. All telephone contacts were audiotaped, and a subset was reviewed during supervision sessions to ensure adherence and to better guide therapists’ intervention strategies. Any deviations from the treatment protocol were brought to the therapist’s attention for remediation.
HOW WELL actual	The number of missed telephone calls from 16 planned calls was comparable between conditions with an average of 1.81 missed calls for FITT‐C and 1.22 for TS. The average call length was slightly longer for the FITT‐C group (37 minutes) compared with TS (30 minutes).

See: Summary of findings 1 Complex information, support and training interventions compared to usual treatment, waiting list or attention control informations for informal caregivers of people with dementia; Summary of findings 2 Complex information, support and training interventions compared to information provision only for informal caregivers of people with dementia

Table 26. TIDieR ‐ Winter 2007

Study	Winter 2007
TIDieR item	Experimental intervention		Control intervention
BRIEF NAME	Telesupport group treatment.		Control condition.
WHY	No described.		No applicable.
WHAT materials	Conference‐calling technology (Toshiba Digital Business Telephone Model DKT2010‐S System and Centrex teleconference service) to link 5 caregivers per group for an hour weekly.		Nurses and an educational booklet.
PROCEDURES	The primary goal was to enhance caregiver ability to manage daily stressors by providing emotional support and validation. Initially, facilitators focused on developing group cohesion. As groups progress, disclosure of intimate problems and personal conflicts emerge. Caregivers express emotions and share coping strategies including cognitive reframing and practical approaches to organizing care routines. They also assist each other in problem solving and share educational resources. The mutual support and validation provided by group members normalize experiences and provide a supportive social network, core to the service model.		No described.
WHO provided	Telesupport groups were conducted by trained social workers.		No applicable.
HOW delivered	Teleconference Group format.		No applicable.
WHERE occurred	At home.
WHEN and HOW MUCH	26 possible sessions in the 6 months following baseline.		No applicable.
TAILORING	Intervention was comprehensively designed, not tailored to cover individual or unmeet needs.	No applicable.
MODIFICATIONS	None described.		No applicable.
HOW WELL planned	No described.		No applicable.
HOW WELL actual	Telesupport group caregivers participated in an aver‐ age of 14.8 sessions, of a possible 26.		No applicable.

Outcome measures

Caregiver burden and caregiver mood were the most frequently reported outcomes across studies.

Caregiver burden was reported in 17 studies using four different scales: the Zarit Burden Inventory (ZBI, versions of 22‐item and 12‐item; 12 studies, 65%); the 24‐item Revised Memory and Behavior Checklist (RMBPC, four studies, 23%); the 10‐item Burden Scale for Family Caregivers (BSFC, one study) and another 6‐item scale (one study).

Caregiver mood was reported in 17 studies using five different scales: the Center for Epidemiologic Studies‐Depression (CES‐D, versions of 20‐item and 10‐item; 12 studies, 71%); the 21‐item Beck Depression Inventory (BDI, two studies, 12%); the 30‐item Geriatric Depression Scale (GDS), the 8‐item Patient Health Questionnaire (PHQ), and the SF‐12 v2 Mental Component Summary (MCS) (one study each). For 16 of these studies, the scale measures depressive symptoms; therefore in the interests of accuracy, we describe the outcome simply as depressive symptoms.

Caregiver HRQoL was reported in four studies with four different scales: the 19‐item Perceived Quality of Life Scale (PQoL), the EuroQol‐5D Visual Analogue Scale (EQ‐5D VAS), the EuroQol‐5D 5‐level (EQ‐5D‐5L), and the 36‐item Short‐Form Health Survey (SF‐36).

Caregiver knowledge and skills was reported in eight studies using six different scales; the Revised Scale for Caregiving Self‐Efficacy (RSCSE with 15‐ and 19‐item versions; three studies, 37%); the 10‐item General Self‐Efficacy Scale (GSES), the 30‐item Alzheimer's Disease Knowledge Scale (ADKS), the 32‐item Trust in Own Abilities (TOA), the 4‐item Caregiver Competence Scale (CCS), and a Visual Analogue Scale (VAS) (one study each).

Care recipient QoL was not reported in any of the included studies.

Because studies used so many different measures with different metrics to report conceptually similar outcomes, we used the SMD for continuous outcomes as the measure of effect size.

Two studies (Duggleby 2018; Huis in het Veld 2020) reported results for caregiver knowledge and skills as mean differences that could not be pooled with the raw effects extracted from the rest of studies. Both studies are included in their respective comparisons and the results reported accordingly.

Funding

All studies but one reported being funded by public grants. Funding source was not stated for Kwok 2013.

Excluded studies

We excluded 52 studies with reasons (see Characteristics of excluded studies).

Risk of bias in included studies

We performed a 'Risk of bias' assessment for each included trial. Our judgements regarding the overall risk of bias in individual studies are presented in Figure 2 and Figure 3.

Figure 2

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Figure 3

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Allocation

All included studies were reported to be randomised. However, many studies did not describe the randomisation procedure in sufficient detail to assess risk of bias, and we considered the risk to be unclear. We asked authors for additional information, which we included, if available, in the 'Risk of bias' tables.

We judged most studies to be at unclear risk of bias for allocation concealment because of insufficient information. We asked authors for additional information, which we included, if available, in the 'Risk of bias' tables.

Blinding

Due to the characteristics of interventions included in this review, blinding of participants and personnel was, in general, not feasible. We rated all studies but Au 2019 and Blom 2015 as being at high risk of bias. Both studies asserted that participants did not know whether the intervention they received was the experimental or the control, and the data were all collected via the Internet with no intermediary interviewer.

We judged most studies to be at high risk of bias for blinding of outcome assessment because most outcomes were subjective and assessed with participant self‐report scales. We considered five studies to be at low risk of bias in this domain (Au 2015; Au 2019; Blom 2015; Huis in het Veld 2020; Tremont 2015).

Incomplete outcome data

We judged most studies to be at low or unclear risk of bias in this domain. We rated the risk as high in 10 studies because they did not provide data about the flow of participants over the study period or presented high attrition rates or differential attrition rates by intervention (Blom 2015; Czaja 2013; Davis 2004; Duggleby 2018; Gustafson 2019; Kajiyama 2013; Mahoney 2003; NCT03417219; Nunez‐Naveira 2016; Tremont 2008).

Selective reporting

About 50% of the studies did not report the availability of a protocol or clinical trial registry entry. We asked authors for additional information, which we included, if available, in the 'Risk of bias' tables. We rated most studies as being at unclear risk of bias for lacking information on prespecification of outcomes at the protocol stage and whether they were primary or secondary outcomes. We judged one study to be at high risk of bias for reporting only statistically significant results (Davis 2004).

Other potential sources of bias

We judged two studies to be at high risk of bias (Au 2015; Winter 2007). The Au 2015 study broke randomisation by adding further data to a pilot study already published; the Winter 2007 study found, after randomisation, a relevant imbalance of prognostic factors at baseline.

Effects of interventions

1 Any intervention involving training, support or both (with or without information) versus usual treatment, waiting list or attention control

Primary outcomes

1.1 Caregiver burden

Based on moderate‐certainty evidence from nine randomised studies including 597 participants, we found that remotely delivered interventions involving training, support or both (with or without information) probably do not affect caregiver burden (SMD ‐0.06, 95% CI ‐0.35 to 0.23; I² = 60%). Negative values indicate less burden in the experimental intervention group. The range of SMD values includes no effect and small effects in favour of either experimental intervention or control. There is no evidence of important differences in effect estimates among types of interventions (Analysis 1.1; Figure 4).

Figure 4

Forest plot of comparison: 1 Any intervention versus usual treatment, waiting list or attention, outcome: 1.1 Caregiver burden.

Support/information or support interventions versus control interventions: based on low‐certainty evidence from three randomised studies including 156 participants, the evidence suggests little or no difference in caregiver burden (SMD 0.12, 95% CI ‐0.43 to 0.66) (Figure 4, upper panel). The range of SMD values includes no effect and moderate effects in favour of either experimental or control interventions.
Training/information or training interventions versus control interventions: based on low‐certainty evidence from three randomised studies including 109 participants, the evidence suggests little or no difference in caregiver burden (SMD ‐0.14, 95% CI ‐0.78 to 0.50) (Figure 4, middle panel). The range of SMD values includes no effect, a large effect in favour of the experimental interventions or a moderate effect in favour of the control interventions.
Interventions including both training and support elements versus control interventions: the evidence from three randomised studies including 332 participants is very uncertain about the effect of interventions including both training and support on caregiver burden (SMD ‐0.17, 95% CI ‐0.70 to 0.36) (Figure 4, lower panel). The range of SMD values includes no effect, a large effect in favour of the experimental interventions and a moderate effect in favour of the control interventions.

1.2 Caregiver depressive symptoms

Based on moderate‐certainty evidence from eight randomised studies including 638 participants, we found that remotely delivered interventions involving training, support or both (with or without information) probably do not affect caregiver depressive symptoms (SMD ‐0.05, 95% CI ‐0.22 to 0.12; I² = 9%). Negative values indicate fewer depressive symptoms in the experimental intervention group. The range of SMD values includes no effect, a small effect in favour of the experimental interventions and a negligible effect in favour of the control interventions. There is no evidence of important differences in effect estimates among types of interventions (Analysis 1.2; Figure 5).

Figure 5

Forest plot of comparison: 1 Any intervention versus usual treatment, waiting list or attention, outcome: 1.2 Caregiver mood.

Support/information or support interventions versus control interventions: based on low‐certainty evidence from four randomised studies including 283 participants, the evidence suggests that the experimental interventions may have little or no effect on caregiver depressive symptoms (SMD ‐0.06, 95% CI ‐0.38 to 0.25) (Figure 5, upper panel). The range of SMD values includes no effect and small effects in favour of either experimental or control interventions.
Training/information or training interventions versus control interventions: based on low‐certainty evidence from one randomised study including 23 participants, the evidence suggests that the experimental interventions may have little or no effect on caregiver depressive symptoms (SMD 0.02, 95% CI ‐0.80 to 0.84) (Figure 5, middle panel). The range of SMD values includes no effect and large effects in favour of either experimental or control interventions.
Interventions involving both training and support elements versus control interventions: based on low‐certainty evidence from three randomised studies including 332 participants, the evidence suggests that the experimental interventions may have little or no effect on caregiver depressive symptoms (SMD ‐0.06, 95% CI ‐0.32 to 0.19) (Figure 5, lower panel). The range of SMD values includes no effect, a small effect in favour of the experimental interventions and a very small effect in favour of the control interventions.

1.3 Caregiver HRQoL

Based on moderate‐certainty evidence from two randomised studies including 311 participants, remotely delivered interventions involving training, support or both (with or without information) probably do not affect caregiver HRQoL (SMD 0.10, 95% CI ‐0.13 to 0.32; I² = 0%). Positive values indicate better HRQoL in the experimental intervention group. The range of SMD values include no effect, a small effect in favour of the experimental interventions and a negligible effect in favour of the control interventions. There is no evidence of important differences in effect estimates among types of interventions (Analysis 1.3; Figure 6).

Figure 6

Forest plot of comparison: 1 Any intervention versus usual treatment, waiting list or attention, outcome: 1.3 Caregiver HRQoL.

Training/information or training interventions versus control interventions: based on low‐certainty evidence from one randomised study including 61 participants, the evidence suggests that there may be little or no difference in caregiver HRQoL (SMD ‐0.12, 95% CI ‐0.62 to 0.38) (Figure 6, upper panel). The range of SMD values includes no effect, a small effect in favour of the experimental interventions and a moderate effect in favour of the control interventions.
Interventions including both training and support elements versus control interventions: based on moderate‐certainty evidence from one randomised study including 250 participants, the experimental intervention probably has little or no effect on caregiver HRQoL (SMD 0.15, 95% CI ‐0.10 to 0.40) (Figure 6, lower panel). The range of SMD values includes no effect, a moderate effect in favour of the experimental intervention and a negligible effect in favour of the control intervention.

1.4 Dropouts for any reason

Based on low‐certainty evidence from eight randomised studies including 661 participants, remotely delivered interventions involving training, support or both (with or without information) probably result in little or no difference in dropouts for any reason (RR 1.15, 95% CI 0.87 to 1.53; I² = 0%). The risk for the experimental intervention group was 223 per 1000 (95% CI 169 to 296) versus 194 per 1000 for the control group. There is no evidence of important differences in risk of dropout among types of interventions (Analysis 1.4; Figure 7).

Figure 7

Forest plot of comparison: 1 Any intervention versus usual treatment, waiting list or attention, outcome: 1.4 Dropouts for any reason.

Support/information or support interventions versus control interventions: based on low‐certainty evidence from three randomised studies including 232 participants, the evidence suggests that the experimental interventions may not affect the risk of dropping out of the study for any reason (RR 1.01, 95% CI 0.59 to 1.73) (Figure 7, upper panel). The risk of dropout in the experimental intervention group was 174 per 1000, versus 172 per 1000 for the control group.
Training/information or training interventions versus control interventions: based on low‐certainty evidence from two randomised studies including 70 participants, the evidence suggests that the experimental interventions may not affect the risk of dropping out for any reason (RR 1.42, 95% CI 0.50 to 4.04) (Figure 7, middle panel). The risk of dropout in the experimental intervention group was 365 per 1000, versus 257 per 1000 for the control group.
Interventions including both training and support elements versus control interventions: based on low‐certainty evidence from three randomised studies including 359 participants, the evidence suggests that the experimental interventions may not affect the risk of dropping out of the study for any reason (RR 1.10, 95% CI 0.75 to 1.63) (Figure 7, lower panel). The risk for intervention was 215 per 1000, versus 195 per 1000 for control.

Secondary outcomes

1.5 Caregiver knowledge and skills

Based on low‐certainty evidence from four randomised studies including 223 participants (Cristancho‐Lacroix 2015; Hattink 2015; Metcalfe 2019; Nunez‐Naveira 2016), the evidence suggests that remotely delivered interventions involving training, support or both (with or without information) may make little or no difference to caregiver knowledge and skills (SMD 0.20, 95% CI ‐0.10 to 0.50; I² = 21%). Positive values favour the experimental intervention group. The range of SMD values includes no effect, a moderate effect in favour of the experimental interventions and a negligible effect in favour of the control interventions. There is no evidence of important differences in effect estimates among types of intervention (Analysis 1.5).

Support/information or support interventions versus control interventions: based on low‐certainty evidence from two randomised studies including 120 participants, the evidence suggests that the experimental interventions may result in little or no difference in caregiver knowledge and skills (SMD 0.18, 95% CI ‐0.18 to 0.54). The range of SMD values includes no effect, a moderate effect in favour of the experimental interventions and a small effect in favour of the control interventions.
Training/information or training interventions versus control interventions: based on low‐certainty evidence from one randomised study including 54 participants, the evidence suggests that the experimental intervention may make little or no difference to caregiver knowledge and skills (SMD ‐0.12, 95% CI ‐0.65 to 0.42). The range of SMD values includes no effect and moderate effects in favour of either the experimental or control intervention.
Interventions including both training and support elements versus control interventions: based on low‐certainty evidence from one randomised study including 49 participants, the evidence suggests that the experimental intervention may result in an increase in caregiver knowledge and skills (SMD 0.61, 95% CI 0.04 to 1.18). The SMD values range from a negligible to a large effect in favour of the experimental intervention.

1.6 Use of health and social resources

Based on moderate‐certainty evidence from one randomised study including 250 participants (Tremont 2015), a remotely delivered intervention involving training, support or both (with or without information) probably results in little or no difference in use of health and social resources (Rate Ratio 1.05, 95% CI 0.93 to 1.19) (Analysis 1.6).

1.7 Admission to institutional care (nursing home placement)

Due to very low‐certainty evidence from one randomised study including 34 participants (Hayden 2012), the evidence is very uncertain about the effect of a remotely delivered interventions involving training, support or both (with or without information) on institutional care (nursing home) placement (RR 0.59, 95% CI 0.11 to 3.11) (Analysis 1.7).

1.8 Care recipient QoL

We found no studies that examined the QoL of the cared‐for person with dementia.

2 Any intervention involving training, support or both (with or without information) versus information alone

Primary outcomes

2.1 Caregiver burden

Based on low‐certainty evidence from nine studies including 650 participants, remotely delivered interventions involving training, support or both (with or without information) may result in a slight reduction in caregiver burden (SMD ‐0.24, 95% CI ‐0.51 to 0.04; I² = 62%) compared to a control condition of information provision alone. Negative values indicate lower burden in the experimental intervention group. The SMD values range from a moderate effect in favour of the experimental interventions to a negligible effect in favour of the control interventions (information provision). There is no evidence of important differences in effect estimates among types of interventions (Analysis 2.1; Figure 8).

Figure 8

Forest plot of comparison: 2 Any intervention versus control information, outcome: 2.1 Caregiver burden.

Support/information or support interventions versus information alone: based on low‐certainty evidence from two randomised studies including 119 participants, the evidence suggests that the experimental interventions may have little or no effect on caregiver burden (SMD ‐0.03, 95% CI ‐0.39 to 0.33) (Figure 8, upper panel). The range of SMD values includes no effect and small effects in favour of either experimental or control interventions.
Training/information or training interventions versus information alone: based on low‐certainty evidence from six randomised studies including 506 participants, the evidence suggests intervention may reduce caregiver burden (SMD ‐0.31, 95% CI ‐0.69 to 0.07) (Figure 8, middle panel). The range of SMD values includes no effect, a large effect in favour of the experimental interventions and a negligible effect in favour of control interventions.
Interventions involving both support and training elements with information versus information alone: based on low‐certainty evidence from one randomised study including 25 participants, the evidence suggests that the interventions may have little or no effect on caregiver burden (SMD ‐0.11, 95% CI ‐0.90 to 0.68) (Figure 8, lower panel). The SMD values range from a large effect in favour of the experimental intervention to a moderate effect in favour of the control intervention.

2.2 Caregiver depressive symptoms

Based on moderate‐certainty evidence from 11 randomised studies including 1100 participants, remotely delivered interventions involving training, support or both (with or without information) probably result in a slight improvement in caregiver depressive symptoms (SMD ‐0.25, 95% CI ‐0.43 to ‐0.06; I² = 53%) compared to a control condition of information alone. Negative values indicate fewer depressive symptoms in the experimental intervention group. SMD values range from a negligible effect to a moderate effect in favour of the experimental interventions. There is no evidence of important differences in effect estimates among types of interventions (Analysis 2.2; Figure 9). We detected no small study effects, that can indicated publication bias (analyses not shown).

Figure 9

Forest plot of comparison: 2 Any intervention versus control information, outcome: 2.2 Caregiver mood.

Support/information or support interventions versus information alone: the evidence from one randomised study including 94 participants is inconclusive about the effect of the experimental intervention on caregiver depressive symptoms (SMD ‐0.28, 95% CI ‐0.69 to 0.13) (Figure 9, upper panel). The range of SMD values includes no effect, a moderate effect in favour of the experimental intervention and a small effect in favour of the control intervention.
Training/information or training interventions versus information alone: based on low‐certainty evidence from eight randomised studies including 816 participants, the evidence suggests the experimental intervention may improve caregiver depressive symptoms (SMD ‐0.31, 95% CI ‐0.54 to ‐0.08) (Figure 9, middle panel). SMD values range from a negligible to a moderate effect in favour of the experimental interventions.
Interventions involving both support and training elements with information versus information alone: based on low‐certainty evidence from two randomised studies including 190 participants, the evidence suggests that the intervention may result in little or no difference to caregiver depressive symptoms (SMD 0.04, 95% CI ‐0.24 to 0.33) (Figure 9, lower panel). The range of SMD values includes no effect and small effects in favour of either experimental or control interventions.

2.3 Caregiver HRQoL

Based on low‐certainty evidence from two randomised studies including 257 participants, remotely delivered interventions involving training, support or both (with or without information) may result in little or no difference in caregiver HRQoL when compared with information alone (SMD ‐0.03, 95% CI ‐0.28 to 0.21; I² = 0%). Positive values indicate better HRQoL in the experimental intervention group. The range of SMD values includes no effect and small effects in favour of either experimental or control interventions (Analysis 2.3; Figure 10).

Figure 10

Forest plot of comparison: 2 Any intervention versus control information, outcome: 2.3 Caregiver HRQoL.

2.4 Dropouts for any reason

Based on moderate‐certainty evidence from 12 randomised studies including 1266 participants, remotely delivered interventions involving training, support or both (with or without information) probably result in an increase in dropouts for any reason (RR 1.51, 95% CI 1.04 to 2.20; I² = 37%). The risk of dropout from the experimental interventions was 196 per 1000 (95% CI 135 to 286) versus 130 per 1000 for the control interventions (information alone). There is no evidence of important differences in risk among types of interventions (Analysis 2.4; Figure 11). We detected no small study effects, which can indicate publication bias (analyses not shown).

Figure 11

Forest plot of comparison: 2 Any intervention versus control information, outcome: 2.4 Dropouts for any reason.

Support/information or support interventions versus information alone: based on low‐certainty evidence from one randomised study including 42 participants, the evidence suggests that the experimental intervention may have little or no effect on dropouts for any reason (RR 1.10, 95% CI 0.17 to 7.09) (Figure 11, upper panel). The risk of dropping out from the experimental interventions was 100 per 1000 versus 91 per 1000 for the control intervention.
Training/information or training interventions versus information alone: based on moderate‐certainty evidence from nine randomised studies including 994 participants, the experimental interventions likely result in a large increase in dropouts for any reason (RR 1.51, 95% CI 0.97 to 2.33) (Figure 11, middle panel). The risk of dropping out from the experimental interventions was 204 per 1000 versus 135 per 1000 for the control interventions.
Interventions involving both support and training elements with information versus information alone: based on very low‐certainty evidence from two randomised studies including 230 participants, the evidence is very uncertain about the effect of the experimental interventions on dropouts for any reason (RR 1.25, 95% CI 0.25 to 6.22) (Figure 11, lower panel). The risk of dropping out from the experimental interventions was 144 per 1000 versus 115 per 1000 for the control interventions.

Secondary outcomes

2.5 Caregiver knowledge and skills

Based on low‐certainty evidence from two randomised studies including 70 participants, the evidence was inconclusive about any effect of remotely delivered interventions involving training, support or both (with or without information) on caregiver knowledge and skills compared to a control condition of information alone (SMD 0.18, 95% CI ‐0.29 to 0.65; I² = 0%). Positive values indicate better caregiver knowledge and skills score in the experimental intervention group. The range of SMD values include no effect, a large effect in favour of the experimental interventions and a moderate effect in favour of the control interventions (Analysis 2.5).

Another two studies (Duggleby 2018; Huis in het Veld 2020), whose results we were unable to include in the meta‐analysis, also reported data on this outcome. Evidence from Duggleby 2018 including 199 participants suggests that the experimental intervention has little or no effect on caregiver knowledge and skills assessed with the 10‐item GSES (MD 0.22, 95% CI ‐0.78 to 1.22). Evidence from Huis in het Veld 2020 including 54 participants also suggests little or no effect of the experimental intervention on caregiver knowledge and skills assessed with the 32‐item TOA (MD ‐2.16, 95% CI ‐6.86 to 2.53).

2.6 Use of health/social resources

We did not find any studies that examined the use of health and social resources.

2.7 Admission to institutional care (nursing home placement)

Based on very low‐certainty evidence from one randomised study including 32 participants (Gant 2007), the evidence is very uncertain about the effect of the experimental intervention on admission to institutional care (nursing home placement) (RR 2.67, 95% CI 0.12 to 60.93) (Analysis 2.6).

2.8 Care recipient QoL

We did not find any studies that examined the QoL of the cared‐for person with dementia.

Discussion

Summary of main results

Our review included 26 studies with a total of 2367 participants. We conducted two comparisons:

Interventions involving training, support or both (with or without information) versus usual treatment, waiting list or attention control;
Interventions involving training, support or both (with or without information) versus information alone.

Interventions involving training, support or both (with or without information) versus usual treatment, waiting list or attention control

Primary outcomes

In this comparison, we found that the experimental interventions probably make little or no difference to caregiver burden (nine studies, 597 participants), caregiver depressive symptoms (eight studies, 638 participants) or caregiver HRQoL (two studies, 311 participants), and probably have little or no effect on dropouts for any reason (eight studies, 661 participants).

Secondary outcomes

We found that the experimental interventions may have little or no effect on caregiver knowledge and skills (four studies, 223 participants), or on use of health and social resources (one study, 250 participants). We found that the evidence is very uncertain about their effect on admission to institutional care (nursing home placement). We did not find studies that examined the QoL of the care recipient.

Interventions involving training, support or both (with or without information) versus information alone

Primary outcomes

In this comparison, we found that interventions including training or support compared to information alone may result in a slight reduction in caregiver burden (nine studies, 650 participants), probably result in a slight reduction in caregiver depressive symptoms (11 studies, 1100 participants), may make little or no difference to caregiver HRQoL (two studies, 257 participants), and probably result in a large increase in dropouts for any reason (12 studies, 1266 participants).

Secondary outcomes

We found ino evidence of an effect of the experimental interventions on caregiver knowledge and skills and very uncertain evidence about the effect on institutional care (nursing home placement) (one study, 32 participants). We did not find studies that examined the use of health and social resources or the QoL of the cared‐for person with dementia.

Overall completeness and applicability of evidence

We believe that we are likely to have identified all the studies that address the efficacy and safety of remotely delivered psychoeducational interventions for caregivers of people with dementia. Our search for evidence was extensive, using the Specialised Register of the CDCIG, as well as running complementary searches in appropriate electronic databases, clinical trials registries, and hand‐searching references from included studies.

The average characteristics of caregivers (mostly females over 60 years of age) and care recipients (83% Alzheimer's diagnosis) are representative of the patient‐caregiver dyads that usually access community health and social care facilities.

There was significant diversity in the outcomes measured in the included trials. Of our primary outcomes, caregiver burden and caregiver mood were most frequently assessed, but assessment of mood in most studies relied on scales of depressive symptoms. Few studies used validated measures of caregiver anxiety, which would also be of interest. Caregivers' HRQoL was assessed in only four trials. We did not find explicit reporting of adverse effects of the interventions and the additional burden they might add to the caregiver's life. Although acceptability is a complex, multi‐dimensional construct, it is difficult to measure and we were only able to approximate it using measures of overall dropout from the studies as a surrogate.

Ten studies reported, if not acceptability, quality control indicators associated with uptake of the interventions, fidelity of implementation or user satisfaction with the programs. As described below, these indicators and assessment areas varied very widely across studies and so we did not attempt to combine them. Au 2019 assessed program fidelity using an unblinding rating system after the end of each session. Brennan 1995 assessed service use by counting service contacts. Cristancho‐Lacroix 2015 assessed the percentage of participants finishing the protocol. Davis 2004 assessed treatment implementation with audiotaped sessions and weekly caregiving logs, among other treatment delivery and receipt checks. Duggleby 2018 assessed if the intervention was used at least once, Martindale‐Adams 2013 assessed percentage of participants attending a minimum number of program sessions. Metcalfe 2019 assessed technology acceptance measuring the perceived usefulness, the perceived ease of use, computer self‐efficacy and behavioural intentions. Nunez‐Naveira 2016 assessed the feasibility of the intervention with three self‐completed questionnaires developed for the study. Torkamani 2014 assessed the web platform with satisfaction questionnaires. Tremont 2015 assessed the intervention with a treatment satisfaction questionnaire.

There are two issues that may compromise the applicability of results from this review: its reliance mostly on populations from high‐income countries with distinct health and social services at community level, and the limitation of the available evidence to the acute effects of interventions, assessed at a median of 16 weeks. Little evidence comes from studies conducted in non‐Western countries (only three studies conducted in China), and most studies were conducted in the United States (15 of 26) which, unlike European countries (eight studies), lacks a universal public health system. We also lack sufficient information to determine whether results apply over longer follow‐up periods, or if effects could be delayed in time, as the developed skills and coping strategies might help caregivers as the dementia progresses. In fact, very few studies (six of 26) presented data to assess effects over time after the end of the experimental interventions (median follow‐up time 24 weeks). These studies describe per protocol results (completers only) rather than intention‐to‐treat results. These missing data limit the applicability of our review.

In summary, evidence is lacking for several important outcomes.The applicability of the evidence from our review may be restricted to populations of high‐income countries, and to the acute effects of interventions.

Quality of the evidence

We used GRADE methods to rate the overall certainty of the evidence for each outcome. All results were downgraded for study limitations. The most widespread risk of bias was the risk of performance bias due to self‐rating of subjective outcomes by participants who were not blind to the treatment. Randomisation methods were not always reported in sufficient detail to allow risk of bias to be assessed. In some studies, there was a risk of attrition bias. We also downgraded some results for inconsistency between studies.

For the comparison of any intervention versus usual treatment, waiting list or attention, we consider the certainty of evidence to be low to moderate. Evidence certainty is low for dropouts for any cause, and moderate for caregiver burden, caregiver mood (depressive symptoms), and caregiver HRQoL (summary of findings Table 1).

For the comparison of any intervention versus control information, we consider the certainty of evidence to be low to moderate. Evidence certainty is low for caregiver burden and caregiver HRQoL, and moderate for caregiver mood (depressive symptoms) and dropouts for any cause (summary of findings Table 2).

Potential biases in the review process

We conducted this systematic review in accordance with Cochrane's current standards. However, the following potential limitations in the review process deserve mention:

The complexity and heterogeneity of the interventions in our review, and their incompletely documented methodology and results, present notable disadvantages when trying to summarise the evidence of complex interventions on caregiver well‐being (Belle 2003; Gaugler 2017). We have used the TIDieR template (Hoffmann 2014) to describe interventions, and following consultation with the Cochrane Dementia and Cognitive Improvement Group, we have classified experimental interventions into three groups for the purposes of subgroup analyses (support alone or with information, training alone or with information, and interventions including both support and training). This was because all but three studies that provided support or training also included provision of information. Further, we have classified control interventions in two categories (one grouping together usual treatment, waiting list and attention control, and the other including information provision only). Pooled analyses for each comparison grouped together all experimental interventions in an overall category labelled 'any treatment'. This inclusive category does not present misclassification bias. However, subgroup analyses by type of experimental intervention may present misclassification issues, which usually tend to bias effects to the null hypothesis of no difference.
This review has examined efficacy at a median time of 16 weeks. There is no information on longer effects of the interventions. Only six studies reported any data for longer follow‐up (median of 24 weeks). However, these studies form a selected subset of all included studies and present data for completers only, which may introduce additional bias because of attrition and missing information. Therefore, we have not conducted a follow‐up analysis.
We contacted study authors of the included trials where the publicly available information for methodological aspects of the studies and/or reported results was unclear or missing. We received responses from some authors (acknowledged in Characteristics of included studies), but many others were impossible to locate or did not respond to our information requests.

An unexpected finding of this review is that, contrary to what might have been anticipated, psychoeducational interventions for caregivers seem to show slightly more efficacy on decreasing burden and improving mood of caregivers,when compared with a control intervention that involved the simple provision of structured information about dementia, than when compared with usual treatment, being on a waiting list, or having non‐specific attention. It is possible that the differential dropout rates we have found in the comparison of experimental interventions with information alone might contribute to this unforeseen result. Since most studies reported outcomes for completers only and experimental interventions were associated with a higher risk of dropout, especially in the studies using an information‐only control, it is conceivable that attrition bias might have inflated effect sizes in favour of interventions in this comparison.

Since most studies have been run in high income countries with developed health and social care systems, it is also possible that usual treatment control groups had access at least to basic information and support as part of standard care services or from other community sources (charities, self‐help groups). If this were the case, usual treatment might represent a level of care that could be efficacious enough to narrow the differences with any other intervention. In support of this argument, it is worth noting that in the first comparison, most studies included usual treatment (n = 6) or attention controls (n = 4), and only two studies (Hattink 2015; Metcalfe 2019) included a waiting list control group.

Agreements and disagreements with other studies or reviews

Other recent reviews have concluded that the efficacy of interventions including information and support or training, is relatively small with absolute average effects of about SMD 0.20 on caregivers' burden ( Abrahams 2018; Williams 2019a), caregivers' mood‐depression (Abrahams 2018), and caregivers' QoL (Abrahams 2018). These effect sizes are slightly higher than those previously reported for remotely delivered interventions (Internet‐based or telephone‐based), but well within the range of variability for effect sizes. Reviews addressing caregiver burden in dementia (Deeken 2019) and dementia among several other chronic conditions (stroke survivors, non‐small cell lung cancer and brain injury) (Ploeg 2018), and reviewing technology‐based and Web‐based interventions respectively, report effect sizes of SMD ‐0.13 (95% CI ‐0.24 to ‐0.02) in favour of any intervention and SMD 0.03 in favour of control (95% CI ‐0.31 to 0.36). Our values overlap with those results: SMD ‐0.06 (95% CI ‐0.35 to 0.23) for comparison versus usual treatment, waiting list or attention control and SMD ‐0.24 (95% CI ‐0.51 to 0.04) for comparison versus control information. For caregiver mood (mainly assessed with depression questionnaires), Deeken 2019 (in dementia caregivers) and Sherifali 2018 (in other chronic condition caregivers) report effect sizes of SMD ‐0.20 (95% CI ‐0.31 to ‐0.10) and SMD ‐0.19 (95% CI ‐0.43 to 0.05) in favour of any intervention, and our values also agree with those results: SMD ‐0.05 (95% CI ‐0.22 to 0.12) for comparison versus usual treatment, waiting list or attention control and SMD ‐0.25 (95% CI ‐0.43 to ‐0.06) for comparison versus control information. For caregiver QoL in chronic conditions (mostly dementia but also stroke, lung cancer and brain injury), Sherifali 2018 reports a negligible effect size of SMD 0.01 in favour of any intervention but with a very wide confidence interval that covers a moderate effect in both directions, in favour and against intervention (95% CI ‐0.49 to 0.51). Our results are also compatible: SMD 0.10 in favour of any intervention (95% CI ‐0.13 to 0.32) for comparison versus usual treatment, waiting list or attention control and SMD ‐0.03 (95% CI ‐0.28 to 0.21) for comparison versus control information. We have not found effect sizes from other reviews dealing with the outcome of dropouts for any reason.

Two qualitative reviews (Egan 2018; Hopwood 2018) indicate positive responses to the use of Internet‐based interventions by caregivers, and that they might be efficacious to improve mental health and reduce caregiver burden. However, they also highlight problems with determining the effectiveness of Internet interventions: the marked methodological diversity across studies (Egan 2018), and the need for high‐quality studies to increase the quality of evidence (Hopwood 2018).

An earlier Cochrane Review that partially overlaps with ours (Lins 2014) found that telephone counselling for informal carers of people with dementia was efficacious for decreasing depressive symptoms (SMD ‐0.32; 95% CI ‐0.63 to ‐0.01) and caregiver burden (SMD ‐0.45; 95% CI ‐0.90 to 0.01). These results are the most favourable we have seen reported for any remotely delivered intervention. The review is restricted to a comparison of telephone counselling with no treatment, usual care, or friendly calls for chatting. It is worth noting that all but one of the studies included in that review are also included in ours. For depressive symptoms, Tremont 2008 and Winter 2007 overlap with ours, as well as Davis 2004, Tremont 2008, and Winter 2007 for caregiver burden. However, we did not include another study that Lins 2014 included for both outcomes (Finkel 2007). Hence, the reviews are not entirely comparable.

In summary, given the large variability of effect sizes reported by other reviews, our results seem consistent with those previously published for caregiver burden, mood and QoL. We cannot compare our results for acceptability of interventions (dropouts for any cause) with other reviews, which did not study this outcome.

Figure 1

Study flow diagram.

Figure 2

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Figure 3

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Figure 4

Forest plot of comparison: 1 Any intervention versus usual treatment, waiting list or attention, outcome: 1.1 Caregiver burden.

Figure 5

Forest plot of comparison: 1 Any intervention versus usual treatment, waiting list or attention, outcome: 1.2 Caregiver mood.

Figure 6

Forest plot of comparison: 1 Any intervention versus usual treatment, waiting list or attention, outcome: 1.3 Caregiver HRQoL.

Figure 7

Forest plot of comparison: 1 Any intervention versus usual treatment, waiting list or attention, outcome: 1.4 Dropouts for any reason.

Figure 8

Forest plot of comparison: 2 Any intervention versus control information, outcome: 2.1 Caregiver burden.

Figure 9

Forest plot of comparison: 2 Any intervention versus control information, outcome: 2.2 Caregiver mood.

Figure 10

Forest plot of comparison: 2 Any intervention versus control information, outcome: 2.3 Caregiver HRQoL.

Figure 11

Forest plot of comparison: 2 Any intervention versus control information, outcome: 2.4 Dropouts for any reason.

Analysis 1.1

Comparison 1: Any intervention versus usual treatment, waiting list or attention, Outcome 1: Caregiver burden

Analysis 1.2

Comparison 1: Any intervention versus usual treatment, waiting list or attention, Outcome 2: Caregiver mood

Analysis 1.3

Comparison 1: Any intervention versus usual treatment, waiting list or attention, Outcome 3: Caregiver HRQoL

Analysis 1.4

Comparison 1: Any intervention versus usual treatment, waiting list or attention, Outcome 4: Dropouts for any reason

Analysis 1.5

Comparison 1: Any intervention versus usual treatment, waiting list or attention, Outcome 5: Caregiver knowledge and skills

Analysis 1.6

Comparison 1: Any intervention versus usual treatment, waiting list or attention, Outcome 6: Use of health and social resources

Analysis 1.7

Comparison 1: Any intervention versus usual treatment, waiting list or attention, Outcome 7: Institutional care ‐ nursing home placement

Analysis 2.1

Comparison 2: Any intervention versus control information, Outcome 1: Caregiver burden

Analysis 2.2

Comparison 2: Any intervention versus control information, Outcome 2: Caregiver mood

Analysis 2.3

Comparison 2: Any intervention versus control information, Outcome 3: Caregiver HRQoL

Analysis 2.4

Comparison 2: Any intervention versus control information, Outcome 4: Dropouts for any reason

Analysis 2.5

Comparison 2: Any intervention versus control information, Outcome 5: Caregiver knowledge and skills

Analysis 2.6

Comparison 2: Any intervention versus control information, Outcome 6: Institutional care ‐ nursing home placement

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
Interventions involving training, support or both (with or without information) compared to usual treatment, waiting list or attention for informal caregivers of people with dementia
Participants or population: Informal caregivers of people with dementia Setting: Community Intervention: Any intervention including information, training or support Comparison: Usual treatment, waiting list or attention control
Outcomes	Risk with usual treatment, waiting list or attention	Risk with any intervention	Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
Caregiver burden assessed with: self‐reported scales ‐ lower score indicates less burden follow up: median 24 weeks	‐	SMD 0.06 lower (0.35 lower to 0.23 higher)	‐	597 (9 RCTs)	⊕⊕⊕⊝ MODERATE ¹	Any intervention probably does not affect caregiver burden. There are no important differences in effect estimates among type of interventions: SMD 0.12 higher for support or support with information SMD 0.14 lower for training or training with information SMD 0.17 lower for multicomponent interventions involving support and training
Caregiver mood assessed with: self‐reported scales ‐ lower score indicates fewer depressive symptoms follow up: median 24 weeks	‐	SMD 0.05 lower (0.22 lower to 0.12 higher)	‐	638 (8 RCTs)	⊕⊕⊕⊝ MODERATE ¹	Any intervention probably does not affect caregiver mood ‐ depressive symptoms. There are no important differences in effect estimates among type of interventions: SMD 0.06 lower for support or support with information SMD 0.02 higher for training or training with information SMD 0.06 lower for multicomponent interventions involving support and training
Caregiver health‐related quality of life (HRQoL) assessed with: self‐reported scales ‐ higher score indicates better QoL follow up: range 6 weeks to 24 weeks	‐	SMD 0.1 higher (0.13 lower to 0.32 higher)	‐	311 (2 RCTs)	⊕⊕⊕⊝ MODERATE ²	Any intervention probably does not affect caregiver health‐related quality of life. There are no important differences in effect estimates among type of interventions: SMD 0.12 lower for training or training with information SMD 0.15 higher for multicomponent interventions involving support and training
Dropouts for any reason assessed with: number of dropouts for any reason throughout the trial follow up: median 18 weeks	Study population		RR 1.15 (0.87 to 1.53)	661 (8 RCTs)	⊕⊕⊝⊝ LOW ^{3 4}	Any intervention probably results in little or no difference in dropouts for any reason. There are no important differences in effect estimates among type of interventions: RR 1.01 for support or support with information RR 1.42 for training or training with information RR 1.10 for multicomponent interventions involving support and training
	194 per 1000	223 per 1000 (169 to 296)	RR 1.15 (0.87 to 1.53)	661 (8 RCTs)	⊕⊕⊝⊝ LOW ^{3 4}
Use of health and social resources assessed with: Frequency of use throughout the trial follow up: mean 24 weeks	Study population		Rate ratio 1.05 (0.93 to 1.19)	4776 person‐weeks (1 RCT)	⊕⊕⊕⊝ MODERATE ⁵	Any intervention probably results in little to no difference in use of health and social resources.
	204 per 1000	214 per 1000 (190 to 243)	Rate ratio 1.05 (0.93 to 1.19)	4776 person‐weeks (1 RCT)	⊕⊕⊕⊝ MODERATE ⁵
Institutional care ‐ nursing home placement assessed with: Frequency of nursing home placement throughout the trial follow up: median 24 weeks	Study population		RR 0.59 (0.11 to 3.11)	34 (1 RCT)	⊕⊝⊝⊝ VERY LOW ^{6 7}	The evidence is very uncertain about the effect of any intervention on institutional care ‐ nursing home placement.
	188 per 1000	111 per 1000 (21 to 583)	RR 0.59 (0.11 to 3.11)	34 (1 RCT)	⊕⊝⊝⊝ VERY LOW ^{6 7}
*The risk in the intervention group (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: Confidence interval; HRQoL: Health‐related quality of life; QoL: Quality of life; RCT: Randomised controlled trial; RR: Risk ratio; SMD: Standardised mean difference
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect
¹Downgraded for study limitations: Presence of high risk for performance and detection bias due to lack of blinding for participants, personnel, and the use of self‐reported outcome scales by unblinded participants. Also, presence of attrition bias. ²Downgraded for study limitations: Presence of high risk for performance bias. ³Downgraded for study limitations: Presence of high risk for performance and detection bias. Lack of blinding for participants might affect the overall rate of dropouts. ⁴Downgraded for imprecision: Few events to get precise estimates (< 300 overall). 95% CI includes a negligible effect and important benefit/harm. ⁵ Downgraded for imprecision: Rate ratio estimated with 1 study only including 250 participants and 4776 person‐weeks of follow up. ⁶Downgraded for study limitations: Unclear risk for selection bias. High risk for performance and detection bias. ⁷Downgraded for imprecision: Only 1 study with very few events (5 overall).

Summary of findings 2. Complex information, support and training interventions compared to information provision only for informal caregivers of people with dementia

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
Interventions involving training, support or both (with or without information) compared to control information alone for informal caregivers of people with dementia
Participants or population: Informal caregivers of people with dementia Setting: Community Intervention: Any intervention including information, training or support Comparison: Information
Outcomes	Risk with control information	Risk with any intervention	Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
Caregiver burden assessed with: self‐reported scales ‐ lower score indicates less burden follow up: median 20 weeks	‐	SMD 0.24 lower (0.51 lower to 0.04 higher)	‐	650 (9 RCTs)	⊕⊕⊝⊝ LOW ^{1 2}	Any intervention may result in a slight reduction in caregiver burden. There are no important differences in effect estimates among type of interventions: SMD 0.03 lower for support or support with information SMD 0.31 lower for training or training with information SMD 0.11 lower for multicomponent interventions involving support and training
Caregiver mood assessed with: self‐reported scales ‐ lower score indicates fewer depressive symptoms follow up: median 20 weeks	‐	SMD 0.25 lower (0.43 lower to 0.06 lower)	‐	1100 (11 RCTs)	⊕⊕⊕⊝ MODERATE ¹	Any intervention probably results in a slight improvement in caregiver mood ‐ depressive symptoms. There are no important differences in effect estimates among type of interventions: SMD 0.28 lower for support or support with information SMD 0.31 lower for training or training with information SMD 0.04 higher for multicomponent interventions involving support and training
Caregiver health‐related quality of life (HRQoL) assessed with: self‐reported scales ‐ higher score indicates better QoL follow up: range 12 weeks to 24 weeks	‐	SMD 0.03 lower (0.28 lower to 0.21 higher)	‐	257 (2 RCTs)	⊕⊕⊝⊝ LOW ^{1 3}	Any intervention may result in little or no difference in caregiver health‐related quality of life.
Dropouts for any reason assessed with: number of dropouts for any reason throughout the trial follow up: median 16 weeks	Study population		RR 1.51 (1.04 to 2.20)	1266 (12 RCTs)	⊕⊕⊕⊝ MODERATE ¹	Any intervention probably results in an increase in dropouts for any reason. There are no important differences in effect estimates among type of interventions: RR 1.10 for support or support with information RR 1.51 for training or training with information RR 1.25 for multicomponent interventions involving support and training
	130 per 1000	196 per 1000 (135 to 286)	RR 1.51 (1.04 to 2.20)	1266 (12 RCTs)	⊕⊕⊕⊝ MODERATE ¹
Use of health and social resources ‐ not reported	‐	‐	‐	‐	‐	No included study reported use of health and social resources.
Institutional care ‐ nursing home placement assessed with: frequency of nursing home placement throughout the trial follow up: mean 12 weeks	Study population		RR 2.67 (0.12 to 60.93)	32 (1 RCT)	⊕⊝⊝⊝ VERY LOW ^{4 5}	The evidence is very uncertain about the effect of any intervention on institutional care ‐ nursing home placement.
	0 per 1000	0 per 1000 (0 to 0)	RR 2.67 (0.12 to 60.93)	32 (1 RCT)	⊕⊝⊝⊝ VERY LOW ^{4 5}
*The risk in the intervention group (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: Confidence interval; HRQoL: Health‐related quality of life; QoL: Quality of life; RCT: Randomised controlled trial; RR: Risk ratio; SMD: Standardised mean difference
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect
¹Downgraded for study limitations: Presence of high risk for performance and detection bias due to the lack of blinding for participants, personnel, and the use of self‐reported outcome scales by unblinded participants. Also, presence of attrition bias. ²Downgraded for inconsistency: Unexplained important heterogeneity (I² > 60% or not overlapping 95% confidence intervals). ³Downgraded for imprecision: Few participants to get precise estimates (< 300 overall). ⁴Downgraded for study limitations: Presence of high risk for performance and detection bias due to lack of blinding for participants, personnel, and the use of self‐reported outcome scales by unblinded participants. ⁵Downgraded for imprecision: Few events to get precise estimates (< 300 overall). 95% CI includes a negligible effect and important benefit/harm.

Summary of findings 2. Complex information, support and training interventions compared to information provision only for informal caregivers of people with dementia

Table 1. TIDieR ‐ Au 2015

Study	Au 2015
TIDieR item	Experimental intervention	Control intervention
BRIEF NAME	Psychoeducation with behavioral activation (PsyED‐BA).	Psychoeducation only (PsyED).
WHY	Based on the Gallagher‐Thompson’s program Coping with Cargiving. The aim is to enhance the use of coping skills via behavioral activation (BA), since it would be more acceptable and helpful than change dysfunctional thoughts, overcome barriers to treatment access and train paraprofessionals to deliver some modules of the program.	General discussion about the psychoeducation program and information packet.
WHAT materials	Caregivers were given a printed copy of the psychoeducation program (adapted from the Chinese version of the Coping with Caregiving manual) together with an information packet with fact sheets concerning local organizations, community resources, and social and mental issues related to dementia. 1.‐ Patient‐Change Workshop, Procedures Manual, Progress Notes, Problem Behavior Tracking forms, written cueing systems for repetitive verbal behavior, audio/videotape or tactile diversion program for repetitive vocal and physical behaviours, and verbal and physical prompting procedures for continence and self‐feeding problems, Caregiver Self‐Change Recording Form. 2.‐ Self‐change Workshop, Problem Behavior Tracking forms, Caregiver‐self change recording form.	Caregivers were given a printed copy of the psychoeducation program (adapted from the Chinese version of the Coping with Caregiving manual) together with an information packet with fact sheets concerning local organizations, community resources, and social and mental issues related to dementia.
PROCEDURES	Participants were contacted by telephone within the same week of the baseline assessment to start the psychoeducation program. There were a total of four weekly intervention telephone calls, each lasting about 30 minutes. ‐ In the first week, all participants were taught about the symptoms and associated behavioral changes of dementia and the possible effects on the caregivers. Participants were also invited to share their caregiving experiences. ‐ In the second week, participants learned about the physical, social, and psychological consequences of stress and being aware of their own stress. They were invited to talk about their own stress. ‐ In the third week, they learned about the principles of identifying and scheduling pleasant events for themselves. ‐ In the fourth week, participants learned about communication: communicating their own needs to seek support from their family members. They also learned about the characteristics of various different types of communications: passive, aggressive, and assertive. Participants had eight biweekly telephone sessions over the following 4 months. Each session consisted of a telephone call lasting 15 to 20 minutes. The first four sessions focused on pleasant event scheduling and the other four sessions on effective communication. The tasks involved in each of the four sessions were as follows: 1. Activity monitoring: how is the participant spending time/ communicating now? 2. Activity scheduling: schedule pleasant event/effective communications 3. Reinforcing or modifying the pleasant event and communication based on feedback or self‐evaluation 4. Activity rescheduling/revision based on changes after modification. Finally, social work services were available upon requests in outpatient departments in which the care recipients received their regular follow‐up.	Participants were contacted by telephone within the same week of the baseline assessment to start the psychoeducation program. There were a total of four weekly intervention telephone calls, each lasting about 30 minutes. Participants had eight biweekly telephone sessions over the following 4 months. Each session consisted of a telephone call lasting 15 to 20 minutes: ‐ Participants were asked to go through the materials of the psychoeducation program and the information package. For each of the telephone session, the participants were asked to select their own topics for general discussion. If the caregiver selected pleasant event scheduling or communications, general discussion would be carried out without any BA procedures. Finally, social work services were available upon requests in outpatient departments in which the care recipients received their regular follow‐up.
WHO provided	Five paraprofessionals recruited trained to administer BA procedures. They were between 55 and 60 years old and had previously completed post‐secondary school training in areas related to human services (i.e., nursing or management). They had subsequently completed a 42‐hour course on Introduction to Psychology and had then received 20 hours of group training led by a social worker and a clinical psychologist (the principal investigator of the study) on BA in the context of the Copy with Caregiving program. Each paraprofessional worker was tested on a mock case before delivering the program to actual caregivers. Ongoing weekly supervision was provided by the clinical psychologist and social worker.
HOW delivered	Individually over the telephone.
WHERE occurred	At home.
WHEN and HOW MUCH	Twelve sessions: sessions 1 to 4 were delivered weekly, sessions 5 to 12 were delivered biweekly. Each session consisted of a telephone call lasting 15 to 20 minutes.
TAILORING	Interventions were comprehensively designed, not tailored to cover individual or unmeet needs.
MODIFICATIONS	None described.
HOW WELL planned	No described.
HOW WELL actual	No described.

Table 1. TIDieR ‐ Au 2015

Table 2. TIDier ‐ Au 2019

Study	Au 2019
TIDieR item	Experimental intervention	Control intervention
BRIEF NAME	Psychoeducation with behavioral activation (TBA).	General Monitoring (TMG).
WHY	Behavioral activation (BA) focuses on constructing reinforcement contingencies that increase functional behavior and self‐efficacy. Self‐efficacy refers to the person’s beliefs about the abilities to exercise control on the events affecting their lives. The intervention was adapted from the Chinese Version of the Coping with Caregiving.	General discussion about the psychoeducation program and information packet.
WHAT materials	Written information, including the forms for pleasant event scheduling, was mailed to participants before the program started.	Information packet.
PROCEDURES	Four weekly sessions of psychoeducation and eight sessions of behavioral activation. Themes of the psychoeducation component were as follows. Session 1 (week 1): · Symptoms and associated behavioral changes in dementia · Stages in dementia · Caregiving roles and demands · Effects on caregivers Session 2 (week 2): · Physical, social and psychological consequence of stress · Identifying stress reactions · Awareness of stress · Stress and well‐being Session 3 (week 3): · The effect of life events on mood · Tracking daily/ weekly events · Identifying pleasant events · Scheduling pleasant events Session 4 (week 4): · Communication needs to family members · Types of communications: passive, aggressive and assertive · Resources available in the community · Planning in the future Themes of the behavioral activation component were: · Session 1 – Review the present use of time. Using the monitoring form · Session 2 – Brain‐storm pleasant events. Scheduling pleasant activities · Session 3 – Review scheduling of events. Discuss how to improve · Session 4 – Review modifications. Consolidate gains on scheduling · Session 5 – Review present social support. Explore new sources of support · Session 6 – Examine communication skills. Explore new options · Session 7 – Review new communications. Discuss how to improve · Session 8 – Review modification. Consolidate gains on support	All TGM participants received four weekly psycho‐education sessions over the phone with the same contents as in the TBA group. These caregivers were then assigned to eight bi‐weekly sessions of general monitoring with no BA intervention. Each of these sessions started with checking in with the caregiver through inviting them to update their caregiving situation. Caregivers were then guided to discuss one of the following topics at each session in this order: 1. caregiver’s health 2. care‐recipient’s needs 3. caregiver’s routines 4. social support. As there were a total of eight sessions, the last four sessions repeated the order of the first four. While some caregivers might report on attempts they made on their own initiative to improve their scheduling and communication, no specific attempt was made to ask them to review these attempts.
WHO provided	An interventionist with a degree in social work delivered all the four sessions of psycho‐education. Six paraprofessional coaches, between 50 and 60 years old and with an undergraduate degree in helping or service professions, delivered the BA or MG interventions, or carried out monitoring. A social worker and a clinical psychologist provided the training and facilitated weekly supervision separately for TBA and TGM coaches.
HOW delivered	Individually over the telephone.
WHERE occurred	At home.
WHEN and HOW MUCH	Twelve sessions: sessions 1 to 4 were delivered weekly and sessions 5 to 12 were delivered biweekly. Each session lasted about 20 minutes.
TAILORING	Interventions were comprehensively designed, not tailored to cover individual or unmeet needs.
MODIFICATIONS	None described.
HOW WELL planned	The program fidelity was assessed by a rating system built into recording form. At the end of each session, all interventionists, including the paraprofessionals, were asked to rate to what extent they were able to follow the protocol for each of the four PE sessions (3 = fully; 2 = adequately with at least 60% of the material covered; 1 = slightly; 0 = not at all). A similar procedure was adopted for each of the 8 sessions for both TBA and TGM. In addition, 10 cases from TBA and 10 cases from TGM were audiotaped. Interventionists’ adherence to the intervention protocol was assessed by two graduate students who had received eight hours of training on the coding scheme. The sessions were coded with reference to four core TBA strategies (activity planning, review to improve on scheduling, develop new help‐seeking communication skills and review to improve on communications) and four core TGM strategies (updating on caregiving situation, health and needs of the caregiver and the care‐recipient, daily routines and family communications).
HOW WELL actual	As planned. Minimal deviations to the main components of the intervention.

Table 2. TIDier ‐ Au 2019

Table 3. TIDieR ‐ Blom 2015

Study	Blom 2015
TIDieR item	Experimental intervention	Control intervention
BRIEF NAME	Mastery of dementia (MoD).	Minimal Intervention (e‐bulletins).
WHY	A guided self‐help Internet course ‘Mastery over Dementia’ (MoD) designed to reduce caregivers’ symptoms of depression and anxiety. MoD included a combination of psycho‐education with active participation of the caregiver, management of behavioral problems, teaching coping strategies, components of cognitive behavioral therapy (cognitive reframing: changing non‐helpful into helpful thoughts), and increasing social support.	No applicable.
WHAT materials	Internet contents.	E‐bulletins (digital newsletters).
PROCEDURES	The Internet course consists of 8 lessons and a booster session with the guidance of a coach monitoring the progress of participants and evaluating the homework. Each lesson has the same structure and consists of information (text material and videos), exercises, and homework, with an evaluation at the start and end of each session. The elements of the course were presented in the following order: · coping with behavioral problems (problem solving) · relaxation · arranging help from others · changing non‐helping thoughts into helping thoughts (cognitive restructuring) · communication with others (assertiveness training). The booster session was provided a month after participants finished the eight lessons, and provided a summary of what was learned. After every lesson, participants sent their homework to a coach via a secure application. The coach sent electronic feedback to caregivers on their homework within three working days. The feedback had to be opened before the next lesson could be started. Participants were automatically reminded to start with a new lesson or to send in their homework if they were not active for a fixed period of time. All participants in this study received feedback from the same coach, a psychologist employed by a health care agency with additional training in cognitive behavioral therapy and experience in the field of dementia.	Caregivers received a minimal intervention consisting of e‐bulletins (digital newsletters) with practical information on providing care for someone with dementia. The bulletins were sent by email according to a fixed schedule (every 3 weeks) over nearly 6 months. The topics of the bulletin, which did not overlap with the content of MoD, were: · driving · holiday breaks · medication · legal affairs · activities throughout the day · help with daily routines · grieving · safety measures in the home · possibilities for peer support. There was no contact with a coach over the length of the study.
WHO provided	Internet intervention. A coach monitored the progression of the participants.	No applicable.
HOW delivered	Online.
WHERE occurred	At home.
WHEN and HOW MUCH	Eight lessons plus a booster session.	The bulletins were sent by email according to a fixed schedule (every 3 weeks) over nearly 6 months.
TAILORING	Interventions were comprehensively designed, not tailored to cover individual or unmeet needs.
MODIFICATIONS	None described.
HOW WELL planned	No described.
HOW WELL actual	No described.

Table 3. TIDieR ‐ Blom 2015

Table 4. TIDieR ‐ Brennan 1995

Study	Brennan 1995
TIDieR item	Experimental intervention	Control intervention
BRIEF NAME	ComputerLink.	None beyond general information to identify local resources.
WHY	The intervention aimed to help caregivers develop problem‐solving skills, manage their own emotions, and increase their knowledge of dementia and caregiver‐support strategies.	No applicable.
WHAT materials	Caregivers received a Wyse 30 terminal and a Everex modem system.	No applicable.
PROCEDURES	ComputerLink provided three functions: · information · decision support · communication The electronic Encyclopedia provided extensive factual information to enhance self‐care and understanding of Alzheimer’s disease and promote health management of the care recipient. The program uses a system to help caregivers to resolve problems basing in the multiattribute utility theory of von Winterfeldt & Edwards. ComputerLink permitted several options for communication among caregivers, including private mail, access to a public bulletin board, and a section of questions and answers. A nurse moderator entered the system daily and read and responded to messages. The nurse served as a moderator and facilitator.	None beyond general information to identify local resources.
WHO provided	Study nurses who had completed an intervention training programme.	No applicable.
HOW delivered	Online.	No applicable.
WHERE occurred	At home.	No applicable.
WHEN and HOW MUCH	Participants had access to the system 24 hours a day during 12 months.	No applicable.
TAILORING	Intervention was comprehensively designed, not tailored to cover individual or unmeet needs.	No applicable.
MODIFICATIONS	None described.	No applicable.
HOW WELL planned	Indicators for logging in the system.	No applicable.
HOW WELL actual	Caregivers logged into the system 3,875 times, They accessed it a mean of 83 times over 12 months. A typical encounter lasted 13 minutes.	No applicable.

Table 4. TIDieR ‐ Brennan 1995

Table 5. TIDieR ‐ Cristancho‐Lacroix 2015

Study	Cristancho‐Lacroix 2015
TIDieR item	Experimental intervention	Control intervention
BRIEF NAME	Diapason.	Treatment as usual.
WHY	The program’s content on the web was based on cognitive theories of stress.	No applicable.
WHAT materials	Web materials.	No applicable.
PROCEDURES	· Session 1: Caregiver stress – this session presents a definition of stress, its causes and consequences on caregivers, risk factors for chronic stress, and mechanisms and effects of relaxation (includes a link to the relaxation training in the Diapason website), as well as strategies for managing stress underlining the importance of looking for respite. · Session 2: Understanding the disease – in this session, the Alzheimer’s disease diagnosis procedure, the symptoms, the progression of the illness, and the consequences on daily life activities for persons with Alzheimer’s disease (PWAD) are explained. · Session 3: Maintaining the loved ones’ autonomy – this session presents the reasons and strategies to involve loved ones in the process of care in order to stimulate the preserved functions and compensate for the lost ones. The session underlines the importance of maintaining the self‐esteem of PWAD. · Session 4: Understanding their reactions – in this session, the most frequent behavioural and psychological symptoms of dementia (BPSD) and their characteristics are succinctly described and illustrated by examples from daily life. The contextual and intrinsic factors that might be associated with them are also described. · Session 5: Coping with behavioral and emotional troubles – this session presents practical advice on how to cope vis‐à‐vis the BPSD described in the previous session. · Session 6: Communicating with loved ones – this session includes the description of the most frequent language troubles and the strategies to modulate and adapt communication to the preserved skills of PWAD. · Session 7: Improving their daily lives – this session presents strategies to facilitate the performance of activities that become difficult or impossible to execute due to apraxia, illustrating them with examples adapted to daily life. · Session 8: Avoiding falls – the session includes practical advice for maintaining and stimulating the relative’s balance and actions to adopt in the event of a fall. In addition, various actions are described to adapt the relative’s home. · Session 9: Pharmacological and non‐pharmacological interventions – this session includes a brief presentation of different interventions available for caregivers in France with pharmacological treatment as well as cognitive and psychological support. · Session 10: Social and financial support – this session presents the different stakeholders and services that may help caregivers in their daily life. The financial and social support provided by the French government is also revised. · Session 11: About the future – this session provides caregivers with information about the role of disease progression anticipation, inviting them to try and foresee solutions keeping a prospective vision, encouraging them to look for further sources of information, and social support to reduce the uncertainty of caregiving situations. · Session 12: In a nutshell – the last session encompasses a summary of the Diapason program, emphasizing the acceptance of support and help and the importance of obtaining more information to anticipate and avoid stressful circumstances. Additionally the website contains other sections that can be consulted at any time: · Relaxation training: guidelines for learning relaxation as well as 2 videos for the modelling of Schultz’s Autogenic Training and Jacobson’s method. · Life Stories: stories about 4 couples, based on testimonials of caregivers, in which difficult situations are illustrated and possible solutions to manage them are discussed (i.e., apathy of patient, caregivers’ isolation). · Glossary: a glossary for technical words (i.e., neuropsychological assessment, aphasia). · Stimulation: practical activities to stimulate autonomy and share pleasant activities with the relatives in daily life. · Forum: a private and anonymous forum to interact with peers, to express their concerns, discuss solutions to daily problems, and share their feelings and experiences. The participants use nicknames to protect their privacy. A clinical psychologist participates in the discussions if necessary (i.e., avoiding aggressive or inappropriate comments).	No applicable.
WHO provided	No web‐moderate self‐provided intervention except when using the forum resource.	No applicable.
HOW delivered	Online.	No applicable.
WHERE occurred	Home.	No applicable.
WHEN and HOW MUCH	The length of the intervention was 3 months, with each weekly session lasting 15 to 30 minutes on average, but there was no time limit and the participants could access different website sections (i.e., relaxation training, forum) for as long as they wished at any time.	No applicable.
TAILORING	Intervention was comprehensively designed, not tailored to cover individual or unmeet needs.	No applicable.
MODIFICATIONS	None described.	No applicable.
HOW WELL planned	Indicators for login in the system.	No applicable.
HOW WELL actual	On average, participants used the website 19.7 times (SD 12.9) and for 262.2 minutes (SD 270.7) during the first 3 months.	No applicable.

Table 5. TIDieR ‐ Cristancho‐Lacroix 2015

Table 6. TIDieR ‐ Czaja 2013

Study	Czaja 2013
TIDieR item	Experimental intervention	Control intervention
BRIEF NAME	Named "intervention condition".	Two control conditions which were combined: attention control and information
WHY	The intervention condition was modelled after the REACH II intervention.	No described.
WHAT materials	A videophone (CISCO IP 7900) was installed in the caregivers’ homes and connected to a Digital Subscriber Line and to a secure server at the host site. Participants in this condition also received a caregiver notebook with basic information about caregiving and community resources.	Caregivers in the attention condition were provided with a notebook with information related to nutrition. The information condition group received educational materials with basic information about dementia, safety, and community resources.
PROCEDURES	The intervention included education and skills training. It was designed to address five caregiver risk areas (safety, social support, problem behaviours, depression, and caregiver health). The intervention was standardized with respect to the treatment domains and strategies used within the treatment domains but individualized with respect to the amount of emphasis within a treatment domain based on a risk assessment administered at baseline. Certified interventionists taught problem solving strategies to deal with the care recipients’ problem behaviours and training on stress management, healthy behavior strategies, community resources, and communication strategies. The intervention included six 1‐hour monthly sessions. The four educational seminars were brief video lectures from experts on topics relevant to caregiving (i.e., update on Alzheimer disease, caregiver depression) and were presented serially; a new video appeared monthly beginning in month 2 of the intervention.The intervention also included five videophone support group sessions, which were interspersed throughout the intervention period. The support groups were structured and included up to six caregivers and a certified group leader. During the support group sessions caregivers received topical information (related to the material covered in the individual intervention sessions) and shared experiences and concerns.	Attention condition: participants received two 1‐hour in‐home sessions and four videophone sessions. They also participated in five telephone support group sessions, which were interspersed with the individual intervention sessions. The support groups followed the same format as those for the intervention group. The content for the attention control condition was structured around nutrition and healthy eating. Information condition: participants were mailed a packet of educational materials and received a brief (<15 minute) telephone “check‐in call” at 3 months post randomisation.
WHO provided	Interventionists. All assessors and interventionists received training and were certified before entering the field.
HOW delivered	Online plus telephone.	Face to face plus telephone.
WHERE occurred	At home.
WHEN and HOW MUCH	6 sessions (1 hour each).	· 6 sessions (1 hour each). · 15 minutes “check‐in‐call”.
TAILORING	Interventions were comprehensively designed, not tailored to cover individual or unmeet needs.
MODIFICATIONS	None described.
HOW WELL planned	Interventionists submitted taped treatment sessions for review and feedback as part of the certification. Treatment implementation was monitored and maintained by weekly supervision meetings. Both interventionists and assessors followed a detailed manual of operations and a delivery assessment form was completed after each contact.
HOW WELL actual	No described.

Table 6. TIDieR ‐ Czaja 2013

Table 7. TIDieR ‐ Davis 2994

Study	Davis 2004
TIDieR item	Experimental intervention	Control intervention
BRIEF NAME	Telephone training or in‐home training (this review only includes evidence from telephone training).	Friendly, socially supportive, telephone calls.
WHY	The study was grounded in the stress, appraisal and coping theory of Lazarus. The premise for the caregivers skill training intervention was that improvements in caregiver's outcomes could be addressed by expanding caregiver's repertoire of effective appraisal and coping strategies through telephone training strategies	No applicable.
WHAT materials	Caregivers received a standardized, loose‐leaf notebook with suggestions for managing frequently encountered Alzheimer’s disease situations. They had to fill out a weekly log on problems they had managed over the past week.	Socially supportive friendly telephone calls.
PROCEDURES	Caregivers participated in training sessions focused on: · general problem solving · caragiver appraisal of behavior problems · written behavioral programs for managing specific problems · strategies for handling effective responses to difficult caregiving situations. They received an initial home visit of 45 minutes to introduce the trained interventionist and to familiarize with the uses of the notebook and log sheets. At each weekly contact, the interventionist reviewed the past week of caregiving problems with the caregiver, discussed the log entries and used selected notebook sections to help the caregiver manage the situation.	Friendly callers made an initial home visit of approximately 45 minutes to introduce themselves. At the time of each contact, callers required about the caregiver's past week as well as any changes in their general health and medication regimen.
WHO provided	Interventionists.
HOW delivered	Telephone.
WHERE occurred	At home.
WHEN and HOW MUCH	12 weekly telephone contacts.
TAILORING	Interventions were comprehensively designed, not tailored to cover individual or unmeet needs.
MODIFICATIONS	No described.
HOW WELL planned	Telephone contacts were tape recorded and reviewed to assess continued adherence.
HOW WELL actual	Telephone contacts averaged 37 minutes (SD=18)	Friendly telephone calls averaged 16 minutes (SD=12).

Table 7. TIDieR ‐ Davis 2994

Table 8. TIDieR ‐ Dowling 2014

Study	Dowling 2014
TIDieR item	Experimental intervention	Control intervention
BRIEF NAME	LEAF (Life Enhancing Activities for Family Caregivers).	Attention control group.
WHY	The positive affect intervention teaches participants a series of behavioral and cognitive “skills” for increasing positive affect coping with health‐related and other life stresses. The skills include noticing and capitalizing on positive events, gratitude, mindfulness, positive reappraisal, focusing on personal strengths, attainable goals, and acts of kindness.	No described.
WHAT materials	No described.
PROCEDURES	Skill‐building Intervention Sessions · Session 1: Positive Events and Gratitude – Identify a positive or meaningful event in the last week and what it means to note, favour, and capitalize on positive events. Homework: write down 3 things that went well each day and why they went well. Begin a gratitude journal (writing one thing each day you are grateful for) that continued throughout the rest of the intervention. · Session 2: Mindfulness – Teach concepts of mindful attention and non‐judgment. Homework: (a) practice awareness of breathing and meditation for 10 minutes daily and (b) once a day take the time to enjoy something that you usually hurry through, do one thing at a time, and pay attention to it. Participants were encouraged to continue the breathing exercise through the remaining weeks of the intervention and to continue the gratitude journal. · Session 3: Positive Reappraisal – Discuss the meaning of positive reappraisal and how to apply it to everyday occurrences. Homework: each day think of one negative or stressful thing that happened. Practice positive reappraisal – why it may not be as bad as initially thought or something good that might come of it. Participants were encouraged to write about their experience reappraising at the end of each day and to continue their gratitude journal and mindful breathing exercises. · Session 4: Personal Strengths and Attainable Goal – Generate a list of personal strengths that can be used in everyday life. Define attainable goals and practice setting one related to self‐care. Homework: achieve attainable goal. Participants were encouraged to write about their goals at end of day and to continue their gratitude journal and mindful breathing exercises. · Session 5: Altruistic Behaviors/Acts of Kindness – Doing for Others – Discuss the positive impact of doing for others. Homework: do something nice for someone else each day. Participants were encouraged to write about their acts of kindness at end of day and to continue their gratitude journal and mindful breathing exercises.	Participants randomised to the control group engaged in 5 one‐on‐one sessions with a facilitator. The sessions were comparable in length to the intervention sessions (approximately 1 hour) but consisted of an interview and did not have any didactic portion or skills practice. Each session began with the completion of the modified Differential Emotions Scale (DES). In addition to these affect questions, each session had qualitative and quantitative questions and activities cantered on a theme (i.e., life history, health history, diet and exercise, social networks, and meaning and spirituality), to keep the sessions different and interesting for participants. Home practice for the control group consisted of the brief daily affect reports. At the start of sessions 2 through 6, the facilitator reviewed the previous week’s affect diary with the participant.
WHO provided	Facilitators for the intervention sessions were trained for content and delivery. Two were clinical nurse specialists in fronto‐temporal dementia and 1 was a psychologist.	The facilitator for the control group sessions was a research associate, familiar with the content of the intervention sessions so as to not engage in the intervention content during control sessions with participants.
HOW delivered	Mainly by video‐conference (only 1 subject participated in‐person, all others participated remotely).
WHERE occurred	At home.
WHEN and HOW MUCH	6 sessions of 1 hour each.
TAILORING	Interventions were comprehensively designed, not tailored to cover individual or unmeet needs.
MODIFICATIONS	None described.
HOW WELL planned	All sessions were audio‐recorded digitally for both quality assurance and intervention evaluation.
HOW WELL actual	No described.

Table 8. TIDieR ‐ Dowling 2014

Table 9. TIDieR ‐ Duggleby 2018

Study	Duggleby 2018
TIDieR item	Experimental intervention	Control intervention
BRIEF NAME	My Tools 4 Care (MT4C).	Information.
WHY	The aim of the intervention, based on transition theory, was to support carers through transitions and increase their self‐efficacy, hope, and health related quality of life. The intervention had multiple components and included choice, as the carers could choose which sections they would like use, and when. User‐generated content is encouraged throughout the intervention as participants may write in sections, add stories, pictures, music etc. Sections include: · About Me · Common Changes to Expect · Frequently Asked Questions and Resources · Important Health Information (about person they are caring for).	No applicable.
WHAT materials	An online version of the Toolkit entitled My Tools 4 Care (MT4C) (https://www.mytools4care.ca/).	A copy of the Alzheimer Society’s booklet The Progression of Alzheimer’s Disease, sent by email.
PROCEDURES	Participants received information for the MT4C website, (i.e., the website address, a unique username and password to access the site for 3 months), and an electronic copy of a Toolkit Checklist in which the participant will be asked to document their use of MT4C (i.e., time spent and content accessed over the 3 months). Participants also received an electronic copy of the Alzheimer Society’s educational booklet, The Progression of Alzheimer’s Disease.	The electronic document consists of a five‐part series that describes the stages of Alzheimer’s disease. It is written for the person with Alzheimer disease, their family and carers and is freely accessible via the Alzheimer Society of Canada website.
WHO provided	Whole intervention was over the Internet.	No applicable.
HOW delivered	Online	Emailed booklet.
WHERE occurred	At home.
WHEN and HOW MUCH	Intervention was accessible for 3 months.	No applicable.
TAILORING	Interventions were comprehensively designed, not tailored to cover individual or unmeet needs.
MODIFICATIONS	None described.	No applicable.
HOW WELL planned	Webpage recorded the time and duration of access.	No applicable.
HOW WELL actual	73% of caregivers used MT4C at least once over the 3‐month period. By 3 months, participants spent most of their time on Section 2 – Common changes to expect (median 15 minutes) and Section 4 – Resources (median 10 minutes).	No applicable.

Table 9. TIDieR ‐ Duggleby 2018

Table 10. TIDieR ‐ Gant 2007

Study	Gant 2007
TIDieR item	Experimental intervention	Control intervention
BRIEF NAME	Video/workbook/telephone coaching condition.	Education/check‐in‐call condition.
WHY	No described.
WHAT materials	Caregivers in the video condition received a 10‐session video series, a workbook from the Dementia Caregiving Skills Program, and weekly telephone calls from a trained coach.	Participants received by mail a 37‐page booklet, Basic Dementia Care Guide and phone calls.
PROCEDURES	The intervention used primarily behavioural strategies: behavioural activation, behavioural management, and stress reduction through relaxation training. The workbook provided didactic and experimental materials to reinforce information presented in each video session. Participants received 12 weekly phone calls by trained research staff who served as coaches. The first 10 calls reinforced each of the video sessions, the last 2 calls served as follow‐up for further application of concepts. Coaches followed a coach manual that provided a script for reviewing didactic materials and assignments with caregivers, and for assisting them in the application of intervention concepts to their problems. During the coaching calls caregivers reported the specific behavioural strategies that they devised, written down, used, and evaluated, based on the behaviour management module that they learned to apply to their situations.	Intervention included information on dementia and suggestions for dealing with a variety of caregiving challenges. In a cover letter, procedures for maximizing the benefits of this educational booklet were provided. Caregivers then received approximately 7 biweekly telephone calls by a trained staff member. In these calls, the staff member checked on the safety of the caregiver and family member, discussed the caregiver’s use of the suggestions from the guide, and responded to questions by referring the caregiver to appropriate sections in the guide. A standardized script was used for calls to participants in this comparison condition.
WHO provided	Coaches following a coach manual.
HOW delivered	By telephone.
WHERE occurred	At home.
WHEN and HOW MUCH	12 weekly telephone calls.	7 bi‐weekly telephone calls.
TAILORING	Interventions were comprehensively designed, not tailored to cover individual or unmeet needs.
MODIFICATIONS	None described.
HOW WELL planned	No described.
HOW WELL actual	No described.

Table 10. TIDieR ‐ Gant 2007

Table 11. TIDieR ‐ Gustafson 2019

Study	Gustafson 2019
TIDieR item	Experimental intervention	Control intervention
BRIEF NAME	D‐CHESS (Dementia–Comprehensive Health Enhancement Support System).	Caregiving booklet.
WHY	D‐CHESS was designed to help with motivation, decision making, stress reduction, and access to services by allowing caregivers to obtain information and support without concern for location, distance, time, confidentiality, or education.	No described.
WHAT materials	Website. Participants were also offered the use of three commercially available sensors: a Bluetooth tracker, a GPS location tracker, and a motion sensor. The Bluetooth tracker would help locate lost items such as keys; the GPS location tracker could be carried by a patient, allowing the caregiver to monitor the patient’s location; and the motion sensor could be used to alert a caregiver if the patient entered specific areas of the home that could be hazardous.	No applicable.
PROCEDURES	This system offered detailed information about dementia, personal accounts of how other caregivers coped, descriptions and locations of services that caregivers may need, and links to vetted online resources. Users could anonymously pose questions to experts, receive suggestions of resources for help, communicate with other caregivers through facilitated discussion groups, and use decision aids to work through key issues. In addition, D‐CHESS services offer other elements proven effective in caregiver programs, such as assertive out‐reach, action planning, and peer and family support, as well as quality informational resources such as expert advice and vetted articles and websites.	No described.
WHO provided	Whole intervention was over the Internet.	No described.
HOW delivered	Online.	Booklet.
WHERE occurred	At home.
WHEN and HOW MUCH	The intervention group received access to the D‐CHESS website for 6 months.	No applicable.
TAILORING	Interventions were comprehensively designed, not tailored to cover individual or unmeet needs.
MODIFICATIONS	None described.
HOW WELL planned	D‐CHESS automatically collected use data, including: number of participants logging on, number of page views, specific services used, discussion group posts, data input to decision aids, and private messages sent to the Alzheimer’s disease information specialist at the Wisconsin Alzheimer’s Disease Research Center.	No applicable.
HOW WELL actual	Overall, participants continued to access D‐CHESS throughout the study, with 100%, 50%, 71%, 71%, 57%, and 64% logging on during months 1 through 6, respectively.	No applicable.

Table 11. TIDieR ‐ Gustafson 2019

Table 12. TIDieR ‐ Hattink 2015

Study	Hattink 2015
TIDieR item	Experimental intervention	Control intervention
BRIEF NAME	STAR E‐Learning (European Skills Training and Reskilling).	Waiting list.
WHY	The STAR training portal was designed to provide dementia care training both for informal and formal caregivers. It is designed as a multilingual e‐learning tool from experts on education, technology, and dementia care.	Participants in the control group were informed they could follow the e‐learning course after post‐test measurements were performed within the experimental intervention.
WHAT materials	The course content was developed from 3 perspectives: · Medical model of dementia: information on types of dementia, symptoms and diagnostics · Functional consequences in daily life and how to compensate for disabilities · Dealing with the psychosocial consequences for the person with dementia and his family	No applicable.
PROCEDURES	The STAR training portal offers the following functionality: · 8 modules on different topics in dementia care (2 at basic level, 6 at intermediate to advanced level): what is dementia; living with dementia; getting a diagnosis; practical difficulties in daily life; emotional impact of dementia; support strategies; positive and empathic communication; looking after yourself · Learning Path Advisor to assess baseline knowledge and confidence, to suggest the point to start the course · Facebook and LinkedIn communities to promote peer support and provide opportunities to contact dementia care professionals The modules consist of text, videos, interactive exercises, knowledge tests, and also include references to other websites, literature, and videos.	No applicable..
WHO provided	Whole intervention was over the Internet.	No applicable.
HOW delivered	Online.	No applicable.
WHERE occurred	At home.
WHEN and HOW MUCH	The intervention group received access to the STAR website for 4 months.	No applicable.
TAILORING	Interventions were comprehensively designed, not tailored to cover individual or unmeet needs. Caregivers can follow their own learning path.
MODIFICATIONS	None described.
HOW WELL planned	Pilot study to assess STAR’s usefulness, user friendliness, and impact on knowledge.	No applicable.
HOW WELL actual	As planned.	No applicable.

Table 12. TIDieR ‐ Hattink 2015

Table 13. TIDieR ‐ Hayden 2012

Study	Hayden 2012
TIDieR item	Experimental intervention		Control intervention
BRIEF NAME	The use of Internet technology for psychoeducation and support with dementia caregivers.		Customary care plus monthly brief telephone calls.
WHY	Data suggest that providing education, social support, and ongoing professional consultation to families involved in the care of a relative with Alzheimer’s disease results in improvement in caregiver psychological status, and sometimes even slows the functional decline of the patient.		No applicable.
WHAT materials	Customary care and access to an intensive, interactive online education and support website.		Customary care and telephone calls.
PROCEDURES	· Online education and support, provided by nine brief educational streaming videos, with brief written brochures, plus non‐compulsory weekly 1‐hour professionally hosted synchronous online chats to review the educational material and problem‐solving care issues, using the rest of the group as sources of feedback. Participants received weekly telephone and email reminders about the chat sessions. · Additional focus cantered on self‐care promotion via behavioural interventions and goal attainment scaling.	Active control condition involving supportive phone calls.
WHO provided	Geriatricians, a geriatric psychiatrist, clinical psychologists, a research sociologist, and master’s and bachelor’s level support staff. Both interventions were provided by the same team.
HOW delivered	Internet.		Telephone.
WHERE occurred	At home.
WHEN and HOW MUCH	6‐month interventions.
TAILORING	Interventions were comprehensively designed, not tailored to cover individual or unmeet needs.
MODIFICATIONS	None described.
HOW WELL planned	No described.
HOW WELL actual	Reports of technological challenges encountered by many of the participants as well as constraints due to VA resources and information security policies.

Table 13. TIDieR ‐ Hayden 2012

Table 14. TIDieR ‐ Huis in het Veld 2020

Study	Huis in het Veld 2020
TIDieR item	Experimental intervention	Control intervention
BRIEF NAME	Major Self‐Management Support Intervention.	· Medium Self‐Management Support Intervention · Minor Self‐Management Support Intervention
WHY	The intervention is based on the 5 steps of the “5A model” of self‐management support and the person‐centered care theory of Kitwood. · assessing the state of behavior, beliefs, and motivation · advising based upon personal health risks · agreeing on a realistic set of goals · assisting in anticipating barriers and developing a specific action plan · arranging follow‐up.	Not described.
WHAT materials	Emails, online videos, e‐bulletins.	Online videos and e‐bulletins.
PROCEDURES	The major intervention arm consists of the following elements: · 3 personal email contacts with a nurse specialized in dementia care · provision of 6 online videos about how to manage behavior changes in a relative with dementia and to improve self‐efficacy in managing with this behavior · provision of e‐bulletins with practical information about different types of behavior changes and how to manage them The nurse gave feedback on assignments and gave feedback on the plan that the family caregiver came up with in the assignments.There were 6 videos about different types of behavior changes that occur frequently (dependent behavior, aggressive behavior, suspicious behavior, apathy or indifference, nighttime restlessness, and masking behavior). Family caregivers chosen the number of videos they watch and the accompanying assignments that they do themselves, depending on their own needs and the behavior changes that occur in their relative with dementia. The behavior changes covered in the bulletins were the same as in the videos. The e‐bulletins involve assignments to help caregivers translate the generic information to their own situation and to reflect on possible causes of the behavior changes, how they want to influence the behavior, and how they want to cope with it.	· The medium self‐management support intervention consists only of the online videos and e‐bulletins · The minor self‐management support intervention consists only of the e‐bulletins, the same as those in the major and medium support interventions
WHO provided	The personal email contacts were handled by a nurse with a Bachelor’s or Master’s qualification in nursing and with follow‐up training in dementia care.
HOW delivered	Email plus online videos plus e‐bulletins.	Online videos plus e‐bulletins.
WHERE occurred	At home.
WHEN and HOW MUCH	6 weeks.
TAILORING	Interventions were comprehensively designed, not tailored to cover individual or unmeet needs.
MODIFICATIONS	None described.
HOW WELL planned	No described.
HOW WELL actual	No described.

Table 14. TIDieR ‐ Huis in het Veld 2020

Table 15. TIDieR ‐ Kajiyama 2013

Study	Kajiyama 2013
TIDieR item	Experimental intervention	Control intervention
BRIEF NAME	iCare Stress Management e‐Training Program (ICC).	Education/Information‐Only Condition (EOC).
WHY	The iCare program is an adaptation of the psycho‐ educational program ‘Coping with Caregiving’ (CwC) developed by Gallagher‐Thompson and colleagues that aims to teach a set of core coping skills to caregivers for stress management. These include: relaxation training; learning to increase everyday pleasant activities (including if possible some pleasant activities with the person with dementia); cognitive restructuring (of unduly negative appraisals regarding caregiving and themselves); and improved communication skills on how to ask for help effectively from family members and relevant community and medical institutions.	The EOC served as a control for the purpose of obtaining new information about dementia and what to expect in caring for dementia patients. It also controlled for the involvement with relevant information on the Internet and for the general cognitive and experiential capacity required to ‘navigate’ in complex Internet materials.
WHAT materials	Website and video‐recordings.
PROCEDURES	Educational modules had a hierarchical structure. Participants were encouraged to practice specific assignments in each module over a 7‐to‐10‐day interval before moving to the next one. The iCare program consists of the following modules: · information segment about what ‘dementia’ means and what are common problems associated with it · components on dealing with stress including techniques for relaxation, stress management, and challenging unduly negative thoughts about caregiving · behavioral activation (increasing everyday positive activities for oneself and the patient with dementia) · communication skills to improve help‐seeking with family and community institutions as well as improving ability to relate to the patient with dementia · managing difficult behaviours of the patient with dementia · review of ‘healthy habits’ (nutrition and exercise) for the caregiver along with information on national resources they can consult for further on‐going assistance At the end of each module, CGs were asked to create their own individual ‘action plans’ in the workbook requiring them to describe what they had learned by completing the module and relevant assignments and how they planned to use the information from that specific module in the future in their daily lives.	Caregivers assigned to the EOC were exposed to a website containing similar navigational features, but with content focused on information about dementia, obtained from reputable national sources. In addition, links to certain video‐taped information were provided.
WHO provided	Online.
HOW delivered	Online.
WHERE occurred	At home.
WHEN and HOW MUCH	During 3 months.	No applicable.
TAILORING	Interventions were comprehensively designed, not tailored to cover individual or unmeet needs.
MODIFICATIONS	None described.
HOW WELL planned	No described.
HOW WELL actual	No described.

Table 15. TIDieR ‐ Kajiyama 2013

Table 16. TIDieR ‐ Kwok 2013

Study	Kwok 2013
TIDieR item	Experimental intervention	Control intervention
BRIEF NAME	Psychoeducational intervention.	DVD information.
WHY	The intervention is based on theories of psychosocial transition and stress coping, under the framework of cognitive behavioral therapy, and focuses on: · providing emotional support · directing caregivers to appropriate resources · encouraging caregivers to attend to their own physical, emotional, and social needs · educating caregivers on strategies to cope with ongoing problems	No described.
WHAT materials	Manualized psychoeducational intervention.	DVD with educational information about dementia caregiving.
PROCEDURES	The participants in the intervention group were educated and given advice on topics related to dementia caregiving: · knowledge of dementia · skills of communicating with the patient · management of behavioral and psychological symptoms of dementia · caregivers’ own emotional issues · resources available in the community · long‐term care plan The topics covered and the schedule of presentation were similar to typical psychoeducation interventions held “on site” at community centres.	No applicable.
WHO provided	A registered social worker.	No applicable.
HOW delivered	By telephone.	DVD.
WHERE occurred	At home.
WHEN and HOW MUCH	12 sessions (approximately 30 minutes per session, one session per week).	No applicable.
TAILORING	Interventions were comprehensively designed, not tailored to cover individual or unmeet needs.
MODIFICATIONS	None described.
HOW WELL planned	No described.
HOW WELL actual	No described.

Table 16. TIDieR ‐ Kwok 2013

Table 17. TIDieR ‐ Mahoney 2003

Study	Mahoney 2003
TIDieR item	Experimental intervention		Control intervention
BRIEF NAME	TLC System.		Control.
WHY	The theoretical model asserts that the background and context of stress (caregiving relationship), plus the stressor (care recipient’s behaviours), as mediated by caregiver mastery, resulted in manifestations of caregiver stress.		No described.
WHAT materials	The technology consisted of an integrated telephone network system and an IVR computer network system.		No applicable.
PROCEDURES	Caregivers dialled in from any standard touch‐tone telephone and heard the narrator greet them by name, review the menu of four module options, and provide the service they requested. The included modules were: · Weekly caregiver’s conversation: monitored the caregiver’s stress levels and provided information on how to manage the care recipient’s behavioural problems. · Personal mailbox: allowed caregivers to send and receive communications through voice mail among themselves or to communicate with a clinical nurse specialist who directly answered or triaged questions to a multidisciplinary professional panel of Alzheimer’s disease experts. · Bulletin board: in‐home telephone support group that provided personal and caregiver group voice mail, similar to a computer chat group. · Activity‐respite conversation: care recipient distraction module designed to reduce disruptive behaviours and to provide caregivers with respite time.		No applicable.
WHO provided	Internet intervention.		No applicable.
HOW delivered	Telephone‐delivered.		No applicable.
WHERE occurred	At home.
WHEN and HOW MUCH	The intervention group had access to the system for a 12‐month period, after which their passwords were disabled.		No applicable.
TAILORING	Intervention was comprehensively designed and not tailored to cover individual or unmeet needs. However, because caregivers have diverse needs and preferences, the intervention offered multiple components with flexibility to effectively appeal to a variety of users. Participants chose the type of component, frequency, duration, and timing of the usage.	No applicable.
MODIFICATIONS	None described.
HOW WELL planned	No described.
HOW WELL actual	No described.

Table 17. TIDieR ‐ Mahoney 2003

Table 18. TIDieR ‐ Martindale‐Adams 2013

Study	Martindale‐Adams 2013
TIDieR item	Experimental intervention	Control intervention
BRIEF NAME	Telephone support (CONNECT).	Printed materials.
WHY	Content and structure of the intervention were based on the 6‐month REACH II intervention of 12 individual in‐home and telephone sessions and five telephone support group sessions.	No described.
WHAT materials	Session materials consisted of a Caregiver Notebook and commercially available pamphlets. The Notebook comprised 29 behaviour – management chapters of five to eight pages each (i.e., bathing, repeated questions), and 17 caregiver stress/coping chapters (i.e., assertiveness, communication, grief) based on research and practice, written in large print and at a fifth‐grade reading level.	Control caregivers received pamphlets on dementia and safety as well as telephone numbers for local resources.
PROCEDURES	Like REACH, the multi‐component intervention targeted caregiving risks, including risks associated with emotional and physical well‐being, safety, burden, social support, and patient behaviour management. The first six sessions covered all Caregiver Notebook chapters that were standardized to include the following topics: · introduction · basic information about dementia and financial and legal issues · safety · caregiver health and well‐being · communication · problem solving The sessions were semi‐structured telephone calls with education, skills‐building, and support. Each session began with a relaxation exercise, caregiver updates, review of strategies tried from the preceding session’s topic, and the group leader’s presentation on a behaviour management or stress and coping topic.The rest of the session included discussion and practice by the entire group on the session’s topic, selection of individual strategies to try, selection of the next session’s topic, and closure, including another signal breath relaxation exercise.	Control caregivers received pamphlets on dementia and safety as well as telephone numbers for local resources.
WHO provided	Trained professionals.
HOW delivered	Telephone.	Printed materials.
WHERE occurred	At home.
WHEN and HOW MUCH	Groups met 14 times over 1 year. The support groups met bi‐weekly for 2 months and monthly thereafter for 1 year, for a total of 14 hour‐long sessions.	No described.
TAILORING	Interventions were comprehensively designed, not tailored to cover individual or unmeet needs.
MODIFICATIONS	None described.
HOW WELL planned	No described.
HOW WELL actual	Of the 77 intervention caregivers, 47 (61%) had at least 75% of the 14 sessions and 59 (76.6%) completed at least half. Six caregivers (7.8%) had fewer than three sessions.	No applicable.

Table 18. TIDieR ‐ Martindale‐Adams 2013

Table 19. TIDieR ‐ Metcalfe 2019

Study	Metcalfe 2019
TIDieR item	Experimental intervention		Control intervention
BRIEF NAME	RHAPSODY.		Wait‐list.
WHY	The intervention adopts a multiple‐component model of dementia carer support inspired by the REACH II Study adapted to the care of people with young onset dementia.		No applicable.
WHAT materials	Online device		No applicable.
PROCEDURES	The programme features seven modules: · What is young onset dementia? · Medical perspective on young onset dementia · Frequent problems and solutions · Dealing with challenging behaviours · Family issues · How to get help · Looking after yourself		No applicable.
WHO provided	Internet intervention.		No applicable.
HOW delivered	Online.		No applicable.
WHERE occurred	At home.
WHEN and HOW MUCH	6 weeks.		No applicable.
TAILORING	Intervention was comprehensively designed, not tailored to cover individual or unmeet needs.	No applicable.
MODIFICATIONS	None described.		No applicable.
HOW WELL planned	No described.
HOW WELL actual	No described.

Table 19. TIDieR ‐ Metcalfe 2019

Table 20. TIDieR ‐ NCT00056316

Study	NCT00056316
TIDieR item	Experimental intervention	Control intervention
BRIEF NAME	Behavioural intervention.	Basic education.
WHY	No described.
WHAT materials	Video, workbook and telephone calls.	37‐page Basic Care Guide and telephone calls.
PROCEDURES	Multicomponent behavioral intervention using 10‐session video series, workbook, and weekly telephone coaching sessions.	Participants receive 37‐page Basic Care Guide and bi‐weekly telephone calls by a trained staff member.
WHO provided	Trained staff member.
HOW delivered	Video plus telephone.	Booklet.
WHERE occurred	No described.
WHEN and HOW MUCH	No described.
TAILORING	No described.
MODIFICATIONS	None described.
HOW WELL planned	No described.
HOW WELL actual	No described.

Table 20. TIDieR ‐ NCT00056316

Table 21. TIDieR ‐ NCT03417219

Study	NCT03417219
TIDieR item	Experimental intervention	Control intervention
BRIEF NAME	Education and Skill Building Rehabilitation‐mobile (ESBR‐m).	Usual care plus supplemental educational materials.
WHY	No described.
WHAT materials	Internet contents.	Educational materials.
PROCEDURES	Participants randomised to the ESBR‐m group will participate in four, 90‐minute group ( 5 participants) sessions. These four sessions are supplemented with a booster session one month following the last intervention session.	Participants randomised to the usual care group will receive supplemental educational materials.
WHO provided	No described.
HOW delivered	Internet.	Booklets.
WHERE occurred	At home.
WHEN and HOW MUCH	Four, 90‐minute group sessions.	No applicable.
TAILORING	No described.
MODIFICATIONS	None described.
HOW WELL planned	No described.
HOW WELL actual	No described.

Table 21. TIDieR ‐ NCT03417219

Table 22. TIDieR ‐ Nunez‐Naveira 2016

Study	Nunez‐Naveira 2016
TIDieR item	Experimental intervention		Control intervention
BRIEF NAME	UnderstAID.		Participants in the control group did not use the application and maintained their usual lifestyle.
WHY	No described.		No applicable.
WHAT materials	Smartphone, PC or Tablet.		No applicable.
PROCEDURES	The understAID consists on a Learning section with a database of contents organized in 5 modules with information about 15 different topics. The topics cover information about the care of a person with dementia and caring for oneself as a caregiver, and consist of text, videos, and images and they also include references to other websites. The modules and topics included in understAID are: · Module 1, Cognitive Declines (Topics: Attention, Memory, and Orientation) · Module 2, Daily Tasks (Topics: Bathing, Incontinence, Massage and Touch, and Physical Exercises) · Module 3, Behavioural Changes (Topics: Anxiety and Agitated Behaviour, Depressive Mood, Manic Symptoms, and Emotional Control and Recognition) · Module 4, Social Activities (Communication and Apathy and Loss of Motivation) · Module 5, You as a Caregiver (Topics: Coping with Own Stress and Motivation). Additionally, it has a Social Network section where the caregivers can interact with other participants and exchange information and opinions. This section was moderated by the researchers of the study.		No applicable.
WHO provided	Internet intervention.		No applicable.
HOW delivered	Online.		No applicable.
WHERE occurred	At home.
WHEN and HOW MUCH	Three months.		No applicable.
TAILORING	Intervention was comprehensively designed, not tailored to cover individual or unmeet needs. However, understAID also included the option of filling in an interactive customisation questionnaire with questions about the level of dementia of the person cared by the informal caregiver and about the preferences, energy, and time availability for learning of the caregiver. By answering these questions at the entry of the application, the information level showed to the informal caregivers is personalized and adjusted to their personal situation.	No applicable.
MODIFICATIONS	None described.		No applicable.
HOW WELL planned	No described.		No applicable.
HOW WELL actual	No described.		No applicable.

Table 22. TIDieR ‐ Nunez‐Naveira 2016

Table 23. TIDieR ‐ Torkamani 2014

Study	Torkamani 2014
TIDieR item	Experimental intervention		Control intervention
BRIEF NAME	ALADDIN.		The participants in the control group were only assessed at the study three time points, without any further contact or intervention.
WHY	No described.		No applicable.
WHAT materials	Internet contents.		No applicable.
PROCEDURES	ALADDIN is a computerized platform designed to offer avenues of support and information to the carer. It also manages and communicates information related to the patient with dementia and their carers from their home to the clinicians, facilitating distant monitoring. ALADDIN has four key features: · ‘ALADDIN TV’ provides information and educational material about dementia, as well as musical entertainment and relaxation and exercise techniques · The ‘SOCIAL NETWORKING’ feature provides a forum for carers using ALADDIN to communicate with each other · ‘MY TASKS’ is the distant monitoring feature of ALADDIN, where carers complete questionnaires about their own and their relatives’ health. Their responses can subsequently generate clinical alerts based on set parameters, comparing new responses to previous data, resulting in the immediate detection of change by the clinicians monitoring the system · The ‘CONTACT US’ feature allows the carer to alert the clinical site and/or generate a request for contact		No applicable.
WHO provided	Internet intervention.		No applicable.
HOW delivered	Online.		No applicable.
WHERE occurred	At home.
WHEN and HOW MUCH	The project ran for 6 months.
TAILORING	Intervention was comprehensively designed, not tailored to cover individual or unmeet needs. However, being the primary users of ALADDIN, the caregivers chose the schedule of their tasks.	No applicable.
MODIFICATIONS	None described.		No applicable.
HOW WELL planned	The system was monitored twice daily by the clinical teams.		No applicable.
HOW WELL actual	No described.		No applicable.

Table 23. TIDieR ‐ Torkamani 2014

Table 24. TIDieR ‐ Tremont 2008

Study	Tremont 2008
TIDieR item	Experimental intervention		Control intervention
BRIEF NAME	Family Intervention: Telephone Tracking – Dementia (FITT‐D).		Standard care.
WHY	The intervention is delivered entirely over the telephone. Theoretical underpinnings of the intervention are based on psychosocial transition, transactional stress and coping, and a systems view of family functioning.		No applicable.
WHAT materials	Treatment manual and interventions guide. Telephone.		No applicable
PROCEDURES	Telephone calls included an initial call (orientation and psychoeducation), weekly calls for six weeks, 12 additional contacts every two weeks, and four monthly termination calls. Initial contacts lasted approximately 60 minutes, and follow‐up contacts lasted about 15–30 minutes. The calls focused on providing emotional support, directing caregivers to appropriate resources, encouraging caregivers to attend to their own physical, emotional and social needs, and teaching caregivers strategies to cope with ongoing problems. The intervention did not provide case management, serve as a question‐and‐answer hotline, or provide psychotherapy over the telephone. The FITT method consists of two stages: · The initial stage, Orientation and Psychoeducation, involved providing caregivers with a rationale for the FITT, an introduction to educational and resource materials, a description of what would happen during future phone contacts and an assessment of key areas thought to be instrumental in addressing caregiver burden and mental health (i.e., caregivers’ health, functioning, mood, thinking and family life). The psychoeducation component of this initial stage involved reviewing information about dementia and common psychological, emotional, psychosocial and medical effects of caregiving · The second stage, Follow‐up, involved weekly and bi‐weekly contacts in which new problems were identified, positive and negative changes in caregivers or care recipients were discussed, and psychoeducational information was reviewed and applied for particular situations. The final four follow‐up calls (monthly) addressed issues of termination by allowing caregivers to anticipate FITT contacts coming to an end and to foster reliance on the support network established during the intervention.		No applicable.
WHO provided	Two master’s level therapists who were trained in the FITT‐D procedure delivered the intervention. Doctoral‐level staff supervised therapists weekly to ensure adherence to the protocol and minimize drift.		No applicable.
HOW delivered	Telephone.		No applicable.
WHERE occurred	At home.
WHEN and HOW MUCH	23 telephone contacts over 12 months. Initial contacts lasted approximately 60 minutes, and follow‐up contacts lasted about 15–30 minutes. The intervention involved approximately 12 hours of contact between the therapist and caregiver.		No applicable.
TAILORING	Intervention was comprehensively designed, not tailored to cover individual or unmeet needs.	No applicable.
MODIFICATIONS	None described.		No applicable.
HOW WELL planned	Sessions were audio‐ taped, and two raters reviewed 30 randomly selected telephone contacts and independently completed adherence and competence scales.		No applicable.
HOW WELL actual	No described.		No applicable.

Table 24. TIDieR ‐ Tremont 2008

Table 25. TIDieR ‐ Tremont 2015

Study	Tremont 2015
TIDieR item	Experimental intervention	Control intervention
BRIEF NAME	Family Intervention: Telephone Tracking‐Caregiver (FITT‐C).	Telephone Support (TS).
WHY	The theoretical framework of FITT‐C is based on psychosocial transition, transactional stress and coping, and a systems view of family functioning. The three underlying theories of FITT‐C are geared toward enhancing coping within the caregiver through active problem solving and facilitating positive changes within the family system.	The control condition was designed to account for nonspecific therapeutic factors, such as interpersonal contact and relationship. The approach was based on a nondirective control condition.
WHAT materials	Telephone.	Telephone.
PROCEDURES	Focused on providing dementia education, emotional support, directing caregivers to appropriate resources, encouraging caregivers to attend to their physical, emotional, and social needs, and teaching caregivers strategies to cope with ongoing problems. The FITT method consisted of an initial orientation and psycho‐ education call that involved providing caregivers with a rationale of FITT, an introduction to the resource materials, a description of future telephone contacts, and education about dementia and effects of caregiving. The remaining contacts identified changes since the last call, assessed key areas for the caregiver (i.e., health, functioning, mood, social support, and family life), and provided interventions and psychoeducation to help caregivers solve problems and use family resources. The focus of the final two calls changed to identifying helpful aspects of contacts and how these functions could be met after termination. After the final call, the therapist prepared a letter briefly highlighting the progress during the intervention and encouraged the caregiver to continue to develop and use adaptive coping strategies.	The primary goal of this condition was to provide nondirective support for caregivers through empathic and reflective listening and open‐ended questioning. The role of the therapist was to provide unconditional positive regard to caregivers and to establish a supportive relationship. Therapists were discouraged from providing directive strategies, such as education, problem‐solving, advice‐giving, or task directives. However, education was not withheld if the caregiver had the wrong information.
WHO provided	Individuals recruited to serve as therapists had experience working with dementia patients and/or care‐ givers or psychotherapy experience working with adults. Therapists were required to be master’s level and received training in dementia and caregiving. FITT‐C therapists received additional training in the intervention method.
HOW delivered	Telephone.
WHERE occurred	At home.
WHEN and HOW MUCH	16 telephone contacts over 6 months. Although initial telephone contacts had standardized durations (approximately 60 minutes), follow‐up contacts varied depending on the severity of caregiver problems (15–30 minutes).
TAILORING	Interventions were comprehensively designed, not tailored to cover individual or unmeet needs.
MODIFICATIONS	None described.
HOW WELL planned	Quality control was implemented by weekly supervision of both the FITT‐C and TS therapists. All telephone contacts were audiotaped, and a subset was reviewed during supervision sessions to ensure adherence and to better guide therapists’ intervention strategies. Any deviations from the treatment protocol were brought to the therapist’s attention for remediation.
HOW WELL actual	The number of missed telephone calls from 16 planned calls was comparable between conditions with an average of 1.81 missed calls for FITT‐C and 1.22 for TS. The average call length was slightly longer for the FITT‐C group (37 minutes) compared with TS (30 minutes).

Table 25. TIDieR ‐ Tremont 2015

Table 26. TIDieR ‐ Winter 2007

Study	Winter 2007
TIDieR item	Experimental intervention		Control intervention
BRIEF NAME	Telesupport group treatment.		Control condition.
WHY	No described.		No applicable.
WHAT materials	Conference‐calling technology (Toshiba Digital Business Telephone Model DKT2010‐S System and Centrex teleconference service) to link 5 caregivers per group for an hour weekly.		Nurses and an educational booklet.
PROCEDURES	The primary goal was to enhance caregiver ability to manage daily stressors by providing emotional support and validation. Initially, facilitators focused on developing group cohesion. As groups progress, disclosure of intimate problems and personal conflicts emerge. Caregivers express emotions and share coping strategies including cognitive reframing and practical approaches to organizing care routines. They also assist each other in problem solving and share educational resources. The mutual support and validation provided by group members normalize experiences and provide a supportive social network, core to the service model.		No described.
WHO provided	Telesupport groups were conducted by trained social workers.		No applicable.
HOW delivered	Teleconference Group format.		No applicable.
WHERE occurred	At home.
WHEN and HOW MUCH	26 possible sessions in the 6 months following baseline.		No applicable.
TAILORING	Intervention was comprehensively designed, not tailored to cover individual or unmeet needs.	No applicable.
MODIFICATIONS	None described.		No applicable.
HOW WELL planned	No described.		No applicable.
HOW WELL actual	Telesupport group caregivers participated in an aver‐ age of 14.8 sessions, of a possible 26.		No applicable.

Table 26. TIDieR ‐ Winter 2007

Comparison 1. Any intervention versus usual treatment, waiting list or attention

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1.1 Caregiver burden Show forest plot	9	597	Std. Mean Difference (IV, Random, 95% CI)	‐0.06 [‐0.35, 0.23]

1.1.1 Support/information plus support	3	156	Std. Mean Difference (IV, Random, 95% CI)	0.12 [‐0.43, 0.66]
1.1.2 Training/information plus training	3	109	Std. Mean Difference (IV, Random, 95% CI)	‐0.14 [‐0.78, 0.50]
1.1.3 Training and support	3	332	Std. Mean Difference (IV, Random, 95% CI)	‐0.17 [‐0.70, 0.36]
1.2 Caregiver mood Show forest plot	8	638	Std. Mean Difference (IV, Random, 95% CI)	‐0.05 [‐0.22, 0.12]

1.2.1 Support/information plus support	4	283	Std. Mean Difference (IV, Random, 95% CI)	‐0.06 [‐0.38, 0.25]
1.2.2 Training/information plus training	1	23	Std. Mean Difference (IV, Random, 95% CI)	0.02 [‐0.80, 0.84]
1.2.3 Training and support	3	332	Std. Mean Difference (IV, Random, 95% CI)	‐0.06 [‐0.32, 0.19]
1.3 Caregiver HRQoL Show forest plot	2	311	Std. Mean Difference (IV, Random, 95% CI)	0.10 [‐0.13, 0.32]

1.3.1 Training/information plus training	1	61	Std. Mean Difference (IV, Random, 95% CI)	‐0.12 [‐0.62, 0.38]
1.3.2 Training and support	1	250	Std. Mean Difference (IV, Random, 95% CI)	0.15 [‐0.10, 0.40]
1.4 Dropouts for any reason Show forest plot	8	661	Risk Ratio (M‐H, Random, 95% CI)	1.15 [0.87, 1.53]

1.4.1 Support/information plus support	3	232	Risk Ratio (M‐H, Random, 95% CI)	1.01 [0.59, 1.73]
1.4.2 Training/information plus training	2	70	Risk Ratio (M‐H, Random, 95% CI)	1.42 [0.50, 4.04]
1.4.3 Training and support	3	359	Risk Ratio (M‐H, Random, 95% CI)	1.10 [0.75, 1.63]
1.5 Caregiver knowledge and skills Show forest plot	4	223	Std. Mean Difference (IV, Random, 95% CI)	0.20 [‐0.10, 0.50]

1.5.1 Support/information plus support	2	120	Std. Mean Difference (IV, Random, 95% CI)	0.18 [‐0.18, 0.54]
1.5.2 Training/information plus training	1	54	Std. Mean Difference (IV, Random, 95% CI)	‐0.12 [‐0.65, 0.42]
1.5.3 Training and support	1	49	Std. Mean Difference (IV, Random, 95% CI)	0.61 [0.04, 1.18]
1.6 Use of health and social resources Show forest plot	1		Rate Ratio (IV, Random, 95% CI)	1.05 [0.93, 1.19]

1.7 Institutional care ‐ nursing home placement Show forest plot	1	34	Risk Ratio (M‐H, Random, 95% CI)	0.59 [0.11, 3.11]

Comparison 1. Any intervention versus usual treatment, waiting list or attention

Comparison 2. Any intervention versus control information

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
2.1 Caregiver burden Show forest plot	9	650	Std. Mean Difference (IV, Random, 95% CI)	‐0.24 [‐0.51, 0.04]

2.1.1 Support/information plus support	2	119	Std. Mean Difference (IV, Random, 95% CI)	‐0.03 [‐0.39, 0.33]
2.1.2 Training/information plus training	6	506	Std. Mean Difference (IV, Random, 95% CI)	‐0.31 [‐0.69, 0.07]
2.1.3 Training and support	1	25	Std. Mean Difference (IV, Random, 95% CI)	‐0.11 [‐0.90, 0.68]
2.2 Caregiver mood Show forest plot	11	1100	Std. Mean Difference (IV, Random, 95% CI)	‐0.25 [‐0.43, ‐0.06]

2.2.1 Support/information plus support	1	94	Std. Mean Difference (IV, Random, 95% CI)	‐0.28 [‐0.69, 0.13]
2.2.2 Training/information plus training	8	816	Std. Mean Difference (IV, Random, 95% CI)	‐0.31 [‐0.54, ‐0.08]
2.2.3 Training and support	2	190	Std. Mean Difference (IV, Random, 95% CI)	0.04 [‐0.24, 0.33]
2.3 Caregiver HRQoL Show forest plot	2	257	Std. Mean Difference (IV, Random, 95% CI)	‐0.03 [‐0.28, 0.21]

2.4 Dropouts for any reason Show forest plot	12	1266	Risk Ratio (M‐H, Random, 95% CI)	1.51 [1.04, 2.20]

2.4.1 Support/information plus support	1	42	Risk Ratio (M‐H, Random, 95% CI)	1.10 [0.17, 7.09]
2.4.2 Training/information plus training	9	994	Risk Ratio (M‐H, Random, 95% CI)	1.51 [0.97, 2.33]
2.4.3 Training and support	2	230	Risk Ratio (M‐H, Random, 95% CI)	1.25 [0.25, 6.22]
2.5 Caregiver knowledge and skills Show forest plot	2	70	Std. Mean Difference (IV, Random, 95% CI)	0.18 [‐0.29, 0.65]

2.5.1 Support/information plus support	1	38	Std. Mean Difference (IV, Random, 95% CI)	0.17 [‐0.47, 0.81]
2.5.2 Training/information plus raining	1	32	Std. Mean Difference (IV, Random, 95% CI)	0.19 [‐0.51, 0.88]
2.6 Institutional care ‐ nursing home placement Show forest plot	1	32	Risk Ratio (M‐H, Random, 95% CI)	2.67 [0.12, 60.93]

Comparison 2. Any intervention versus control information