Mental practice for treating upper extremity deficits in individuals with hemiparesis after stroke

Ruth E Barclay; Ted J Stevenson; William Poluha; Brenda Semenko; Julie Schubert

doi:10.1002/14651858.CD005950.pub5

. 2020 May 25;2020(5):CD005950. doi: 10.1002/14651858.CD005950.pub5

Mental practice for treating upper extremity deficits in individuals with hemiparesis after stroke

Ruth E Barclay ^1,^✉, Ted J Stevenson ², William Poluha ³, Brenda Semenko ⁴, Julie Schubert ⁵

Editor: Cochrane Stroke Group

PMCID: PMC7387111 PMID: 32449959

Abstract

Background

Stroke is caused by the interruption of blood flow to the brain (ischemic stroke) or the rupture of blood vessels within the brain (hemorrhagic stroke) and may lead to changes in perception, cognition, mood, speech, health‐related quality of life, and function, such as difficulty walking and using the arm. Activity limitations (decreased function) of the upper extremity are a common finding for individuals living with stroke. Mental practice (MP) is a training method that uses cognitive rehearsal of activities to improve performance of those activities.

Objectives

To determine whether MP improves outcomes of upper extremity rehabilitation for individuals living with the effects of stroke.

In particular, we sought to (1) determine the effects of MP on upper extremity activity, upper extremity impairment, activities of daily living, health‐related quality of life, economic costs, and adverse effects; and (2) explore whether effects differed according to (a) the time post stroke at which MP was delivered, (b) the dose of MP provided, or (c) the type of comparison performed.

Search methods

We last searched the Cochrane Stroke Group Trials Register on September 17, 2019. On September 3, 2019, we searched the Cochrane Central Register of Controlled Trials (the Cochrane Library), MEDLINE, Embase, the Cumulative Index to Nursing and Allied Health Literature (CINAHL), PsycINFO, Scopus, Web of Science, the Physiotherapy Evidence Database (PEDro), and REHABDATA. On October 2, 2019, we searched ClinicalTrials.gov and the World Health Organization International Clinical Trials Registry Platform. We reviewed the reference lists of included studies.

Selection criteria

We included randomized controlled trials (RCTs) of adult participants with stroke who had deficits in upper extremity function (called upper extremity activity).

Data collection and analysis

Two review authors screened titles and abstracts of the citations produced by the literature search and excluded obviously irrelevant studies. We obtained the full text of all remaining studies, and both review authors then independently selected trials for inclusion. We combined studies when the review produced a minimum of two trials employing a particular intervention strategy and a common outcome. We considered the primary outcome to be the ability of the arm to be used for appropriate tasks, called upper extremity activity. Secondary outcomes included upper extremity impairment (such as quality of movement, range of motion, tone, presence of synergistic movement), activities of daily living (ADLs), health‐related quality of life (HRQL), economic costs, and adverse events. We assessed risk of bias in the included studies and applied GRADE to assess the certainty of the evidence. We completed subgroup analyses for time since stroke, dosage of MP, type of comparison, and type of arm activity outcome measure.

Main results

We included 25 studies involving 676 participants from nine countries. For the comparison of MP in addition to other treatment versus the other treatment, MP in combination with other treatment appears more effective in improving upper extremity activity than the other treatment without MP (standardized mean difference [SMD] 0.66, 95% confidence interval [CI] 0.39 to 0.94; I² = 39%; 15 studies; 397 participants); the GRADE certainty of evidence score was moderate based on risk of bias for the upper extremity activity outcome. For upper extremity impairment, results were as follows: SMD 0.59, 95% CI 0.30 to 0.87; I² = 43%; 15 studies; 397 participants, with a GRADE score of moderate, based on risk of bias. For ADLs, results were as follows: SMD 0.08, 95% CI ‐0.24 to 0.39; I² = 0%; 4 studies; 157 participants; the GRADE score was low due to risk of bias and small sample size. For the comparison of MP versus conventional treatment, the only outcome with available data to combine (3 studies; 50 participants) was upper extremity impairment (SMD 0.34, 95% CI ‐0.33 to 1.00; I² = 21%); GRADE for the impairment outcome in this comparison was low due to risk of bias and small sample size. Subgroup analyses of time post stroke, dosage of MP, or comparison type for the MP in combination with other rehabilitation treatment versus the other treatment comparison showed no differences. The secondary outcome of health‐related quality of life was reported in only one study, and no study noted the outcomes of economic costs and adverse events.

Authors' conclusions

Moderate‐certainty evidence shows that MP in addition to other treatment versus the other treatment appears to be beneficial in improving upper extremity activity. Moderate‐certainty evidence also shows that MP in addition to other treatment versus the other treatment appears to be beneficial in improving upper extremity impairment after stroke. Low‐certainty evidence suggests that ADLs may not be improved with MP in addition to other treatment versus the other treatment. Low‐certainty evidence also suggests that MP versus conventional treatment may not improve upper extremity impairment. Further study is required to evaluate effects of MP on time post stroke, the volume of MP required to affect outcomes, and whether improvement is maintained over the long term.

Plain language summary

Mental practice to improve arm function and arm movement in individuals with hemiparesis after stroke

Review question

Does mental practice improve the outcomes of upper extremity rehabilitation for individuals living with the effects of stroke?

Background

Mental practice is a process through which an individual repeatedly mentally rehearses an action or task without actually physically performing the action or task. The goal of mental practice is to improve performance of those actions or tasks. Mental practice has been proposed as a potential adjunct to physical practice that is commonly performed by survivors of stroke undergoing rehabilitation.

Search date

We searched 10 electronic databases and two clinical trials databases in September 2019.

Study characteristics

We included 25 studies with 676 study participants. Thirty‐four per cent of participants were women. In all studies, participants were randomly allocated to groups. The studies, which were reported from nine countries, measured one or more of the following outcomes: arm function for real‐life tasks appropriate to the upper limb (e.g. drinking from a cup, manipulating a doorknob), the amount and quality of movement in the arm, and activities of daily living. We sought but did not find evidence related to health‐related quality of life, economic costs, and adverse events.

Key results

Our review of the available literature provided moderate‐certainty evidence that mental practice, when added to other physical rehabilitation treatment, produced improved outcomes compared to use of the other rehabilitation treatment alone. Evidence to date shows improvements in arm function and arm movement. It is not clear whether (1) mental practice added to physical practice produces improvements in activities of daily living; (2) mental practice alone compared to conventional treatment is beneficial in improving motor control of the arm; (3) how much mental practice could produce the best results; (4) and whether mental practice is best used at a particular time after stroke. No adverse effects or harms were reported in any of the studies.

Certainty of the evidence

For mental practice added to other physical rehabilitation treatment compared to use of the other rehabilitation treatment alone, the certainty of evidence was moderate for arm function and arm movement outcomes based on some challenges in study design. The certainty of evidence was low for the activities of daily living outcome based on study design and the small number of participants included. For the mental practice compared to conventional treatment comparison, the certainty of evidence for the arm movement outcome was determined to be low for the same reasons.

Summary of findings

Background

Description of the condition

Stroke is caused by the interruption of flow of blood to the brain (ischemic stroke) or the rupture of blood vessels within the brain (hemorrhagic stroke) (Heart and Stroke 2018). As of 2016, stroke was the second leading cause of death worldwide, causing an estimated 5.5 million deaths (Gorelick 2019). People who survive a stroke may undergo sudden and intense changes in perception, cognition, mood, speech, health‐related quality of life, and function (e.g. difficulty walking or using the arm) (Mayo 1999). It is estimated that stroke expenditures account for 3% to 4% of all healthcare expenditures in Western countries (Struijs 2006).

This review uses terminology from the International Classification of Functioning, Disability and Health (ICF). The ICF defines activity as "the execution of a task or action by an individual", while activity limitations are defined as "difficulties an individual may have in executing activities" (World Health Organization 2001). In the ICF, body functions are "the physiological functions of body systems (including psychological function)" and body structures are "anatomical parts of the body such as organs, limbs, and their components" (World Health Organization 2001). Impairments are "problems in body function or structure such as significant deviation or loss" (World Health Organization 2001).

Activity limitations of the upper extremity (arm and hand) are a common finding for individuals living with the effects of stroke, with prevalence reported as between 33% and 95% of this population (Nakayama 1994; Parker 1986; Sunderland 1989). These activity limitations may occur because of impairments caused by the stroke in motor ability (Andrews 2003; Canning 2004; Chae 2002; Mercier 2004), somatosensation (Carey 1995), and perceptual ability (Chen Sea 1993; Katz 1999). This has led to a variety of interventions that rehabilitation professionals may use to maximize the upper extremity function (activity) of the individuals they treat (Corbetta 2015; French 2016; Mehrholz 2018; Pollock 2014; Rose 2004; Thieme 2018; Woodford 2007). A feature of these interventions that is thought to be critical to their success is the number of repetitions the individuals perform. Conventional rehabilitation programs have been criticized for failing to provide the opportunity to perform a sufficient volume of practice. Moreover, study authors have suggested that increasing the amount of time spent in therapeutic activity in rehabilitation units would improve the units' effectiveness as learning environments, thereby improving a patient's recovery (Bernhardt 2008).

Description of the intervention

Mental practice (MP) is a training method that calls for cognitive rehearsal of activities, with repetition of the internal representation of a movement, for the explicit purpose of improving performance of those activities. The movement is not actually produced but is, instead, created in the individual's imagination (Jackson 2001; Page 2001b). Practically speaking, a person can be said to have participated in MP when he or she adheres to a set of imagined task performances (e.g. picking up a cup) or movements (e.g. reaching out with his or her arm), visualized from first or third person perspective, for a defined number of imagined repetitions or a defined amount of time. These visualizations are performed without external visual cueing (e.g. watching performance on a videotape), although training in the procedure may use this modality.

Consensus indicates that, in neurologically intact individuals, MP is more effective than no practice in producing skilled movement; however, physical practice whereby goal movements are produced seems to be superior to MP alone (Di Rienzo 2014; Driskell 1994; Felz 1983). It has been suggested that tasks requiring problem‐solving are better suited to and would benefit more from MP (Batson 2004).

How the intervention might work

Utilizing MP may lead to neuroplasticity, based on activity imagined and practiced (Di Rienzo 2014). Studies of neuroimaging have shown that similar overlapping brain areas are activated both in mental imagery and with physical movement. Areas activated include pre‐motor, primary motor, primary somatosensory, and parietal areas.(Di Rienzo 2014).

Why it is important to do this review

We completed a systematic review of randomized controlled trials available as of November 2009 that examined effects of MP on the recovery of various indicators of upper extremity activity and impairment in individuals living with the effects of a stroke (Barclay‐Goddard 2011). At that time, we found limited evidence from only six studies (total sample size = 119 participants) to suggest that MP, when combined with other rehabilitation treatment, appears to be beneficial in improving upper extremity activity after stroke compared with other rehabilitation treatment without MP. The strength of our conclusions was, obviously, tempered by the small number of suitable trials. However, our literature search also revealed six additional ongoing studies, anticipated to recruit a total of 485 participants that seemed relevant to our review. If these trials were completed, our analysis could be strengthened considerably. Therefore, an updated review should be undertaken to account for these additional trials.

Objectives

To determine whether MP improves outcomes of upper extremity rehabilitation for individuals living with the effects of stroke.

Methods

Criteria for considering studies for this review

Types of studies

We sought parallel‐group randomized controlled trials (RCTs) and cross‐over trials for this review.

Types of participants

We sought studies examining adult survivors of stroke who have deficits in upper extremity activity. We used a clinical definition of stroke. Selection of studies was not influenced by the chronicity or severity of the stroke nor by the root cause of the activity limitation (e.g. motor versus sensory versus perceptual deficits).

Types of interventions

We looked for trials that incorporated MP of upper extremity movements or tasks. The MP could be a stand‐alone intervention or could serve as an adjunct to a conventional intervention. The definition of a conventional intervention varies in the literature; conventional intervention may be described as usual care or as usual physiotherapy or occupational therapy treatment. We sought trials that conducted the following comparisons.

MP alone versus no intervention.
MP alone versus conventional intervention.
MP alone versus placebo MP (e.g. counting backward by sevens, listening to information).
MP with conventional intervention versus conventional intervention alone.
MP with conventional intervention versus placebo mental activity with conventional intervention.
MP plus other therapeutic intervention versus other therapeutic intervention alone.

Types of outcome measures

Primary outcomes

Primary outcomes focused on activity and activity limitations of the upper extremity. The upper extremity includes the arm and hand. Upper extremity activity outcomes include indicators of the ability to perform an upper extremity function, such as grasping a cup or folding a towel. Some examples of upper extremity activity outcome measures include:

Box and Block test (Mathiowetz 1985);
Action Research Arm Test (Hsieh 1998);
Wolf Motor Function Test (Wolf 2001); and
Motor Activity Log (Uswatte 2006).

We also looked to include hand function tests such as the Jebsen Test of Hand Function (Jebsen 1969). We reviewed all outcomes and planned to combine only those that measured the same concepts.

Secondary outcomes

Secondary outcomes focused on changes in body structure or function (i.e. impairment) and included indicators of motor recovery or quality of movement (amount and pattern of movement) of the upper extremity. Impairment outcome measures may include items that reflect quality of movement, incorporating aspects of range of motion, strength, tone, or synergistic movement of the upper extremity. Examples of upper extremity impairment standardized outcome measures include:

Fugl‐Meyer Test of Sensorimotor Ability (Fugl‐Meyer 1980); and
Impairment Inventory of the Chedoke‐McMaster Stroke Assessment (Gowland 1993).

We sought other secondary outcomes as well, including:

activities of daily living, such as the Barthel Index (BI) or the Functional Independence Measure (FIM) (Keith 1987; Mahoney 1965);
health‐related quality of life, for example, the Stroke Impact Scale (Duncan 1999);
indicators of the economic costs of interventions; and
any measurement of adverse effects of interventions, including death.

Search methods for identification of studies

See the methods for the Cochrane Stroke Group Specialised register. We searched for trials in all languages and arranged for the translation of trials when necessary.

Electronic searches

The last search of databases for the original review was conducted in November 2009. New searches were begun as of October 2009, allowing one month of overlap.

We searched the Cochrane Stroke Group Trials Register, which was last searched by the Stroke Group Information Specialist on 17 September 2019. In addition, we searched the following bibliographic databases: the Cochrane Central Register of Controlled Trials (CENTRAL) (the Cochrane Library; 3 September 2019) (Appendix 1); MEDLINE Ovid (1946 to 3 September 2019) (Appendix 2); Embase Ovid (1980 to 3 September 2019) (Appendix 3); the Cumulative Index to Nursing and Allied Health Literature (CINAHL) (EbscoHost) (1982 to 3 September 2019) (Appendix 4); PsycINFO ProQuest (1872 to 3 September 2019) (Appendix 5); Scopus (1996 to 3 September 2019) (Appendix 6); and Web of Science, Core Collection (1955 to 3 September 2019) (Appendix 7). We also searched the Physiotherapy Evidence Database (PEDro) (www.pedro.org.au/) (3 September 2019) and the specialist rehabilitation research database REHABDATA (www.naric.com) (3 September 2019).

Searching other resources

In an effort to identify further published, unpublished, and ongoing trials, we:

reviewed the reference lists of included studies; and
searched the following ongoing trials registers (2 October 2019):
- ClinicalTrials.gov (www.clinicaltrials.gov) (Appendix 8); and
- World Health Organization International Clinical Trials Registry Platform (apps.who.int/trialsearch) (Appendix 9).

Data collection and analysis

Selection of studies

Two review authors (RB,TS) independently screened the titles and abstracts of citations produced by the literature search, using Covidence software (www.covidence.org/), and excluded irrelevant studies. We obtained the complete text of all remaining citations and reviewed the texts in further detail, selecting studies that appeared to meet the inclusion criteria. Two review authors reviewed each article; four review authors were involved in the process (RB, TS, BS, JS). Review authors who reviewed the studies are rehabilitation professionals (physiotherapists and occupational therapists) with experience in stroke rehabilitation and are familiar with the background material regarding MP. We arranged for one trial in Portugese to be translated by two physiotherapists. We resolved any disagreement through negotiation, with a third review author serving as arbiter as necessary.

Data extraction and management

Two review authors extracted data from each study, also through Covidence. In total, four review authors took part in this step (RB, JS, TS, BS). We resolved differences as described above. We extracted information regarding numbers of participants, sex, age, time since stroke, side of lesion, dosage of MP, outcome measures, interventions, and type of study. This information is included in the Characteristics of included studies tables. Data extraction of trial results consisted of immediate post‐intervention assessment along with any follow‐up assessment findings. We carried out an analysis when the review produced a minimum of two trials employing a particular intervention strategy and examining a common outcome. We used the Cochrane Review Manager software for all data analyses (RevMan 2014).

Assessment of risk of bias in included studies

Two review authors extracted data and examined the quality of each included trial using the Cochrane "Risk of bias" tool (Higgins 2011). In total, four review authors took part in this step (RB, JS, TS, BS). We resolved differences as described above.

We assessed selection bias (sequence generation and allocation concealment), performance bias (blinding of participants and blinding of personnel), detection bias (blinding of outcome assessment), attrition bias (incomplete outcome data), and reporting bias (selective outcome reporting) (Higgins 2011). We assessed each potential bias as high risk, low risk, or unclear risk by reviewing each research article. We chose high risk if the specific bias appeared to be present, low risk if bias did not appear to be present, and unclear risk if there was no mention of the bias or if bias was not clear in the article. We used the "Risk of bias" tool judgements as described in the Cochrane Handbook for Systematic Reviews of Interventions, Table 8.5a (Higgins 2011).

Measures of treatment effect

All outcomes produced continuous data. We calculated the mean difference (MD) and the 95% confidence interval (CI) when identical measures were combined. We used the standardized mean difference (SMD) and the 95% CI for different outcomes that measured the same constructs.

Unit of analysis issues

With cross‐over RCTs, we used only the first period for analysis.

Dealing with missing data

If data required for meta‐analysis were missing or were provided in a format that could not be used for analysis, we contacted the study author to request missing values, or we estimated values if possible.

Assessment of heterogeneity

We analyzed trials for statistical heterogeneity using the I² statistic, which describes the percentage of variability in effect estimates due to heterogeneity.

Assessment of reporting biases

We reviewed funnel plots for the primary comparison and outcomes.

Data synthesis

We carried out analyses using the random‐effects model.

Subgroup analysis and investigation of heterogeneity

We completed preplanned subgroup analyses based on time since stroke of six months or longer post stroke and less than six months post stroke, as well as a subgroup analysis on the dosage of MP (total time of MP over the course of the study), with less than or equal to 360 minutes versus greater than 360 minutes of total MP time. Based on the studies included, we also completed post hoc subgroup analyses of observed or self‐perceived upper extremity activity outcomes and types of comparisons (conventional treatment or placebo and conventional).

Sensitivity analysis

We planned to complete a sensitivity analysis by excluding studies in which allocation concealment was at high or unclear risk of bias. However, of the 25 included studies, none had high risk of bias and four were at low risk of bias; these four were not combined (Ietswaart 2011; Nilsen 2012; Siqueira 2013; Thara 2015).

When heterogeneity was above 50%, we completed a sensitivity analysis by excluding studies that included constraint‐induced movement therapy (CIMT) (Kim 2017; Page 2009; Park 2015b), as these could potentially be different. We also completed a sensitivity analysis by excluding the study in which all participants had experienced neglect in addition to the stroke (Welfringer 2011).

We also undertook post hoc sensitivity analyses of studies for which random sequence allocation was unclear and blinding of outcome assessors was at high or unclear risk of bias.

GRADE and "Summary of findings"

We created "Summary of findings" tables for each comparison that we analyzed and included the following outcomes: upper extremity activity, upper extremity impairment, activities of daily living, quality of life, economic costs, and adverse events. We used the five GRADE considerations of study limitations, inconsistency of results, indirectness of evidence, imprecision, and publication bias to assess the quality of the body of evidence for each outcome, considering the studies included in the meta‐analyses (Schünemann 2013). We used methods described in the GRADE Handbook (Schünemann 2013), and we created the "Summary of findings" table using GRADEpro GDT (GRADEpro GDT). We explained all decisions to downgrade the quality of studies by using footnotes.

Results

Description of studies

Results of the search

Our 2011 review retrieved 1100 articles, of which we reviewed 29 in full (Barclay‐Goddard 2011). We initially included 10 of the 29 articles; however, on closer inspection, we decided to exclude four of them (BovendEerdt 2010; Hemmen 2007; Liu 2004; Miltner 1999), leaving six studies with a total of 119 participants for inclusion (Müller 2007; Page 2001a; Page 2005; Page 2007; Page 2009; Riccio 2010). Of the six studies included, we could combine data from a total of 102 participants in five studies (Page 2001a; Page 2005; Page 2007; Page 2009; Riccio 2010).

Through the current search, we identified 1768 records and one additional record by reviewing reference lists of included studies. We removed 781 duplicate citations and then 881 at the title‐abstract level of review, leaving 107 for full‐text review. Seventy‐four of these did not meet our inclusion criteria. Eight were ongoing studies and six were not reported in English and we were unable to translate them. One trial was translated from Portuguese to English (Siqueira 2013). This produced 19 additional studies (Alves 2018; Ietswaart 2011; Kanwar 2011; Kim 2015; Kim 2017; Li 2018; Nayeem 2012; Nilsen 2012; Oh 2016; Page 2000; Page 2011; Park 2015a; Park 2015b; Siqueira 2013; Sun 2013; Thara 2015; Timmermans 2013; Wang 2019; Welfringer 2011).

In total, 25 studies are included in this review. See the PRISMA diagram, which summarizes numbers from the 2011 review and the 2020 review (Figure 1).

Included studies

All included studies used random allocation to form treatment groups. Two used a cross‐over design (Oh 2016; Riccio 2010). Six studies formed three groups (Alves 2018; Ietswaart 2011; Kanwar 2011; Müller 2007; Nilsen 2012; Siqueira 2013), and one study formed four groups (Page 2011). All the other studies formed two groups.

One study presented data as slopes of mean change (Müller 2007), one presented change scores (Riccio 2010), one did not present in the format of mean and standard deviation (SD) (Kim 2017), and two others did not present SDs for all outcomes (Page 2005; Page 2011). We contacted the authors of these five studies for further information, obtaining data that could be used for the analysis from Müller 2007, Riccio 2010, Page 2005 (for the 2011 review), and Kim 2017 (for the 2020 review). We did not receive information on SDs for the Page 2011 trial; therefore we did not include it in the meta‐analysis.

For Kanwar 2011 and Li 2018, we estimated mean and SD values from graphs provided in the papers. For Timmermans 2013, we converted median and interquartile range values to an estimate of mean and SD. For Nilsen 2012, we estimated means and SEs from the bar graph figures and then converted SEs to SDs. For all cases, two review authors performed the estimates independently to ensure accuracy of the process. Nayeem 2012 presented median and SD and used median as an approximation of the mean. Oh 2016 presented individual scores ‐ we calculated mean and SD for post‐treatment scores before the cross‐over.

In total, 676 participants were included in the 25 studies; 424 participants were men, 220 were women, and 32 were not reported. Five studies were performed in the Republic of Korea, seven in the USA, three in China, three in India, two in Germany, two in Brazil, and one each in Netherlands, Italy, and Scotland. Details of the 25 included studies can be found in the Characteristics of included studies tables.

The experimental group for most (21/25) studies paired MP with a conventional treatment (Ietswaart 2011; Kanwar 2011; Kim 2015; Kim 2017; Li 2018; Nayeem 2012; Oh 2016; Page 2000; Page 2001a; Page 2005; Page 2007; Page 2009; Page 2011; Park 2015a; Park 2015b; Riccio 2010; Sun 2013; Thara 2015; Timmermans 2013; Wang 2019; Welfringer 2011), and four administered MP alone (Alves 2018; Müller 2007; Nilsen 2012; Siqueira 2013). All groups in Müller 2007 had Bobath and proprioceptive neuromuscular facilitation (PNF) treatment for their impairments (other than hand) after stroke. We considered CIMT a type of "conventional" therapy, as it entails primarily physical practice of tasks with increased practice time. The content of MP was arm and hand movements or activities (e.g. ADLs) for 23 studies; the MP studied by Li 2018 and Müller 2007 focused on hand movements only.

Comparison groups included conventional treatment (Alves 2018; Ietswaart 2011; Kanwar 2011; Kim 2017; Li 2018; Müller 2007; Nayeem 2012; Oh 2016; Page 2009; Park 2015a; Park 2015b; Riccio 2010; Siqueira 2013; Sun 2013; Wang 2019; Welfringer 2011); placebo (relaxation or listening to information or music) plus conventional (Ietswaart 2011; Page 2000; Page 2001a; Page 2005; Page 2007; Page 2011); third person MP (Nilsen 2012); relaxation (placebo) (Nilsen 2012); passive kinesiotherapy (Siqueira 2013); MP plus motor plus video (Kanwar 2011); conventional plus neurodevelopmental treatment (NDT) (also considered conventional) (Timmermans 2013); various lengths of MP (Page 2011); mirror therapy and home exercise (Thara 2015); and virtual reality (Alves 2018).

Treatment intensity (dosage) varied across studies in terms of frequency of treatment sessions, length of sessions, and total number of sessions. Dosage was categorized as greater than 360 minutes of mental practice (Alves 2018; Ietswaart 2011; Li 2018; Müller 2007; Nilsen 2012; Page 2009; Page 2011; Park 2015b; Riccio 2010; Siqueira 2013; Sun 2013; Thara 2015; Timmermans 2013; Wang 2019; Welfringer 2011), or as less than or equal to 360 minutes of mental practice (Kanwar 2011; Kim 2015; Kim 2017; Nayeem 2012; Oh 2016; Page 2000; Page 2001a; Page 2005; Page 2007; Park 2015a).

Time since stroke was categorized as greater than or equal to six months post stroke (Alves 2018; Kim 2015; Kim 2017; Nayeem 2012; Nilsen 2012; Page 2000; Page 2001a; Page 2005; Page 2007; Page 2009; Park 2015a; Park 2015b), less than six months post stroke (Ietswaart 2011; Li 2018; Müller 2007; Riccio 2010; Siqueira 2013; Sun 2013; Thara 2015; Timmermans 2013; Wang 2019; Welfringer 2011), or not stated (Kanwar 2011; Oh 2016; Page 2011).

Instruments used to reflect our primary outcome (upper extremity activity) were the Action Research Arm Test (ARAT) (Ietswaart 2011; Li 2018; Page 2001a; Page 2005; Page 2007; Page 2009; Page 2011; Park 2015a; Park 2015b; Thara 2015; Welfringer 2011), the Arm Function Test (Riccio 2010), individual items from the Jebsen Hand Function Test (Kim 2017; Müller 2007; Nilsen 2012), the Wolf Motor Function Test (WMFT) (Kim 2015), and the Motor Activity Log‐Quality of Movement (MAL‐QOM) (Kanwar 2011; Kim 2017; Nayeem 2012; Oh 2016; Page 2005). The Arm Function Test was not clearly described (Riccio 2010).

Instruments reflecting upper extremity impairment level outcomes were the Fugl‐Meyer Upper Extremity (FM) (Alves 2018; Kim 2015; Li 2018; Nayeem 2012; Nilsen 2012; Oh 2016; Page 2000; Page 2001a; Page 2007; Page 2009; Page 2011; Park 2015a; Park 2015b; Siqueira 2013; Sun 2013; Timmermans 2013; Wang 2019), the Motricity Index (Riccio 2010), grip/pinch strength (Ietswaart 2011; Müller 2007), and accelerometry or three‐dimensional (3D) motion analysis (Kim 2017; Timmermans 2013).

Activities of daily living (ADLs) were assessed in four studies with the Barthel Index or the modified Barthel Index (BI) (Ietswaart 2011; Park 2015a; Park 2015b; Timmermans 2013), as well as the Functional Independence Measure (FIM) (Siqueira 2013). A health‐related quality of life measure was noted in one study, which used the Modified Functional Limitation Profile (Ietswaart 2011). The Frenchay Activities Index was used in one study as a measure of instrumental ADLs (Timmermans 2013), and the Canadian Occupational Performance Measure was used in another study (Nilsen 2012).

We did not find indicators of the economic costs of interventions nor any measurement of adverse effects of interventions, including death.

Excluded studies

We excluded four studies and reported the reasons for their exclusion in Characteristics of excluded studies.

Ongoing studies

We identified eight ongoing studies relevant to this review (ChiCTR‐IOR‐16008137; CTRI/2016/05/006930; CTRI/2019/01/016966; KCT0002574; NCT00379392; NCT01651533; NCT03251209; SLCTR/2017/031). Details of these studies are available in the Characteristics of ongoing studies table.

Studies awaiting classification

Six studies were written in Chinese; we are not able to translate them at this time. We have classified these studies as "awaiting classification" (Fu 2010; Hu 2010; Liu 2015; Liu 2016; Ran 2013; Tang 2014).

Risk of bias in included studies

See Figure 2 (Risk of bias graph) and Figure 3 (Risk of bias summary table).

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

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

Allocation

Four studies appear to have low risk of selection bias related to allocation concealment (Ietswaart 2011; Nilsen 2012; Siqueira 2013; Thara 2015). The other 21 articles had unclear risk of selection bias ‐ typically, allocation concealment was not reported.

All studies were described as randomized; however, 12 studies did not describe how participants were randomized, making risk of selection bias unclear as related to randomization (Alves 2018; Kanwar 2011; Kim 2015; Kim 2017; Li 2018; Müller 2007; Nayeem 2012; Nilsen 2012; Page 2000; Park 2015a; Sun 2013; Wang 2019).

Blinding

Blinding of both participants and personnel did not occur in most studies. Due to the nature of the intervention, this would have been difficult. In some studies with a placebo, participants were likely blinded, but not necessarily the therapist as well. We deemed 16 studies to have high risk of performance bias (Alves 2018; Ietswaart 2011; Kanwar 2011; Kim 2015; Kim 2017; Li 2018; Müller 2007; Nayeem 2012; Nilsen 2012; Page 2005; Page 2007; Riccio 2010; Siqueira 2013; Sun 2013; Thara 2015; Wang 2019). Page 2001a appeared to have low risk of performance bias. For the other eight studies, risk was unclear.

Most studies described blinded assessors, leading to low risk of detection bias (Alves 2018; Ietswaart 2011; Nilsen 2012; Page 2001a; Page 2005; Page 2007; Page 2009; Page 2011; Park 2015b; Riccio 2010; Siqueira 2013; Sun 2013; Timmermans 2013; Wang 2019; Welfringer 2011). Risk of detection bias appeared to be high in Kim 2017. For the remaining seven studies, risk was unclear.

Incomplete outcome data

Sixteen studies appeared to have low risk of attrition bias; for five studies, risk was unclear (Page 2000; Park 2015a; Riccio 2010; Siqueira 2013; Thara 2015), and one study had high risk of bias (Kim 2017).

Selective reporting

Risk of reporting bias appeared to be low in all studies.

Other potential sources of bias

We considered publication bias. We evaluated funnel plots for Analysis 1 (Analysis 1.1, Analysis 1.2, Analysis 1.3) ‐ mental practice in addition to other treatment versus other treatment (± placebo). For the outcome of upper extremity activity, the funnel plot appears close to symmetrical (Figure 4). For the outcome of upper extremity impairment, the graph does appear slightly asymmetrical (Figure 5). For activities of daily living (ADL), only four studies reported this outcome (Figure 6). Asymmetry could be due to publication bias but could also be due to other reasons, such as heterogeneity (Higgins 2019). For both upper extremity activity and upper extremity impairment outcomes, overall heterogeneity was moderate at I² = 39% and 43%, respectively. For ADLs, I² was 0%.

1.1 — Comparison 1: Mental practice in addition to other treatment vs other treatment (± placebo), Outcome 1: Upper extremity activity

1.2 — Comparison 1: Mental practice in addition to other treatment vs other treatment (± placebo), Outcome 2: Upper extremity impairment

1.3 — Comparison 1: Mental practice in addition to other treatment vs other treatment (± placebo), Outcome 3: Activities of daily living

Funnel plot of comparison: 1 mental practice in addition to other treatment versus other treatment (± placebo), outcome: 1.1 Upper extremity activity (arm function).

Funnel plot of comparison: 1 mental practice in addition to other treatment versus other treatment (± placebo), outcome: 1.2 Upper extremity impairment (motor control): FM, Motricity Index, grip strength.

Funnel plot of comparison: 1 mental practice in addition to other treatment versus other treatment (± placebo), outcome: 1.3 Activities of daily living: Barthel, modified Barthel.

Effects of interventions

See: Table 1; Table 2

Summary of findings 1. Mental practice in addition to other treatment compared to other treatment (± placebo) for treating upper extremity deficits in individuals with hemiparesis after stroke.

Mental practice in addition to other treatment compared to other treatment (±placebo) for treating upper extremity deficits in individuals with hemiparesis after stroke
Patient or population: treating upper extremity deficits in individuals with hemiparesis after stroke Setting: clinic, home, research laboratory, or unclear Intervention: mental practice in addition to other treatment Comparison: other treatment (± placebo)
Outcomes	*Anticipated absolute effects (95% CI)**		Relative effect (95% CI)	No. of participants (studies)	Certainty of the evidence (GRADE)	Comments
Outcomes	Risk with other treatment (± placebo)	Risk with mental practice in addition to other treatment	Relative effect (95% CI)	No. of participants (studies)	Certainty of the evidence (GRADE)	Comments
Mean upper extremity activity	0	SMD 0.66 higher (0.39 higher to 0.94 higher)	‐	397 (15 RCTs)	⊕⊕⊕⊝ Moderate^a
Mean upper extremity impairment	0	SMD 0.59 higher (0.3 higher to 0.87 higher)	‐	397 (15 RCTs)	⊕⊕⊕⊝ Moderate^a
Mean activities of daily living	0	SMD 0.08 higher (0.24 lower to 0.39 higher)	‐	157 (4 RCTs)	⊕⊕⊝⊝ Low^a,b
Mean quality of Life	‐	See comment	‐	(0 studies)	‐	Could not do meta‐analysis: only 1 study had a quality of life outcome
Mean economic costs: not reported	‐	‐	‐	‐	‐	Could not do meta‐analysis: no studies had an economic cost outcome.
Mean adverse events: not reported	‐	‐	‐	‐	‐	No adverse events were described
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; RCT: randomized controlled trial; SMD: standardized 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.

Mental practice compared to conventional therapy for treating upper extremity deficits in individuals with hemiparesis after stroke
Patient or population: treating upper extremity deficits in individuals with hemiparesis after stroke Setting: clinic or research laboratory Intervention: mental practice Comparison: conventional
Outcomes	*Anticipated absolute effects (95% CI)**		Relative effect (95% CI)	No. of participants (studies)	Certainty of the evidence (GRADE)	Comments
Outcomes	Risk with conventional therapy	Risk with mental practice	Relative effect (95% CI)	No. of participants (studies)	Certainty of the evidence (GRADE)	Comments
Mean upper extremity activity	‐	See comment	‐	(0 studies)	‐	Could not do meta‐analysis: only 1 study had an upper extremity activity outcome
Mean upper extremity impairment	0	SMD 0.34 higher (0.33 lower to 1 higher)	‐	50 (3 RCTs)	⊕⊕⊝⊝ Low^a,b
Mean activities of daily living	‐	See comment	‐	(0 studies)	‐	Could not do meta‐analysis: only 1 study had an activities of daily living outcome
Mean quality of life: not reported	‐	‐	‐	‐	‐	Could not do meta‐analysis: no study had a quality of life outcome
Mean economic costs: not reported	‐	‐	‐	‐	‐	Could not do meta‐analysis: no study had an economic cost outcome
Mean adverse events: not reported	‐	‐	‐	‐	‐	No adverse events were described
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; RCT: randomized controlled trial; SMD: standardized 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.

Date	Event	Description
3 September 2019	New search has been performed	We updated the searches. We have included 19 new studies since the 2011 review with 557 new participants. A total of 25 studies with 676 participants are included
3 September 2019	New citation required and conclusions have changed	Main conclusion has not changed, but we found more evidence, evaluated a new outcome, and added new subgroups

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1.1 Upper extremity activity	15	397	Std. Mean Difference (IV, Random, 95% CI)	0.66 [0.39, 0.94]
1.1.1 Observed: ARAT, WMFT, arm functional test	11	323	Std. Mean Difference (IV, Random, 95% CI)	0.61 [0.28, 0.94]
1.1.2 Self‐perceived: MAL‐QOM	4	74	Std. Mean Difference (IV, Random, 95% CI)	0.89 [0.41, 1.38]
1.2 Upper extremity impairment	15	397	Std. Mean Difference (IV, Random, 95% CI)	0.59 [0.30, 0.87]
1.3 Activities of daily living	4	157	Std. Mean Difference (IV, Random, 95% CI)	0.08 [‐0.24, 0.39]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
2.1 Upper extremity activity	13	367	Std. Mean Difference (IV, Random, 95% CI)	0.63 [0.34, 0.92]
2.1.1 Greater than or equal to 6 months post stroke	8	179	Std. Mean Difference (IV, Random, 95% CI)	0.75 [0.44, 1.06]
2.1.2 Less than 6 months post stroke	5	188	Std. Mean Difference (IV, Random, 95% CI)	0.48 [‐0.04, 0.99]
2.2 Upper extremity impairment	14	387	Std. Mean Difference (IV, Random, 95% CI)	0.56 [0.27, 0.85]
2.2.1 Greater than or equal to 6 months post stroke	8	180	Std. Mean Difference (IV, Random, 95% CI)	0.60 [0.16, 1.05]
2.2.2 Less than 6 months post stroke	6	207	Std. Mean Difference (IV, Random, 95% CI)	0.53 [0.13, 0.94]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
3.1 Upper extremity activity	15	397	Std. Mean Difference (IV, Random, 95% CI)	0.66 [0.39, 0.94]
3.1.1 Greater than 360 minutes MP	6	214	Std. Mean Difference (IV, Random, 95% CI)	0.59 [0.09, 1.09]
3.1.2 Less than or equal to 360 minutes MP	9	183	Std. Mean Difference (IV, Random, 95% CI)	0.72 [0.42, 1.03]
3.2 Upper extremity impairment	15	397	Std. Mean Difference (IV, Random, 95% CI)	0.59 [0.30, 0.87]
3.2.1 Greater than 360 minutes MP	8	243	Std. Mean Difference (IV, Random, 95% CI)	0.62 [0.24, 1.00]
3.2.2 Less than or equal to 360 minutes MP	7	154	Std. Mean Difference (IV, Random, 95% CI)	0.56 [0.08, 1.04]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
4.1 Upper extremity activity	15	397	Std. Mean Difference (IV, Random, 95% CI)	0.66 [0.39, 0.94]
4.1.1 Mental practice plus conventional vs placebo plus conventional	5	140	Std. Mean Difference (IV, Random, 95% CI)	0.71 [0.06, 1.35]
4.1.2 Mental practice plus conventional vs conventional only	10	257	Std. Mean Difference (IV, Random, 95% CI)	0.69 [0.40, 0.98]
4.2 Upper extremity impairment	15	397	Std. Mean Difference (IV, Random, 95% CI)	0.59 [0.30, 0.87]
4.2.1 Mental practice plus conventional vs placebo plus conventional	4	131	Std. Mean Difference (IV, Random, 95% CI)	0.28 [‐0.07, 0.63]
4.2.2 Mental practice plus conventional vs conventional only	11	266	Std. Mean Difference (IV, Random, 95% CI)	0.70 [0.32, 1.07]

*Study characteristics*
Methods	Study design: RCT Study grouping: parallel group Number of groups: 3
Participants	Baseline characteristics Experimental: mental imagery age: 55.53 years (10.80) sex: 5 women, 10 men stroke etiology: 13 ischemic, 2 hemorrhagic side of lesion: R 8, L 7 Experimental: virtual reality age: 55.05 years (11.52) sex: 6 women, 11 men stroke etiology: 15 ischemic, 2 hemorrhagic side of lesion: R 9, L 8 Control: conventional therapy age: 60.16 years (13.19) sex: 3 women, 7 men stroke etiology: 6 ischemic, 4 hemorrhagic side of lesion: R7, L3 Included criteria: stroke; 6 months or longer post stroke, hemiparesis, able to walk, ages between 30 and 80 years, minimum MMSE score based on education (not stated) Excluded criteria: unable to walk, cognitive or communication impairment, serious cardiovascular disease, aphasia, apraxia, hemineglect or hemiplegia
Interventions	Intervention characteristics Experimental: mental imagery frequency: 2 weeks, 5 days/week × 75 minutes each components: patients watched videos of transferring beans with a spoon, pronation and supination of clay on a fork, stacking cones. Patients closed eyes and imagined practicing the tasks 10 times. Patients performed activities on verbal commands number of minutes total: 750 minutes Experimental: virtual reality frequency: 2 weeks, 5 days/week × 75 minutes each components: Nintendo Wii sports: passive stretching of UE; instructions on use ‐ controller in affected hand with elastic if necessary; playing 3 games of 15 minutes each ‐ baseball, tennis, boxing; cool‐down number of minutes total: 750 minutes Control: conventional therapy frequency: not recorded components:continued with conventional therapy ‐ not recorded number of minutes total: not recorded
Outcomes	FMA‐UE outcome type: continuous outcome
Identification	Sponsorship source: none noted Country: Brazil Setting: physical therapy clinic Author's name: Andreia Maria Silva Vilela Terra Institution: Federal University of Alfenas Email: andreiamarias96@gmail.com Address: Neurofunctional Physical Therapy Laboratory, Federal University of Alfenas, Alfenas, Minas Gerais, Brazil
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Judgement comment: randomized, but method not stated
Allocation concealment (selection bias)	Unclear risk	Judgement comment: not stated
Blinding of participants and personnel (performance bias) All outcomes	High risk	Quote: "the examiners responsible for the interventions and participants were not blinded for the interventions due to the specificity of each protocol"
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote: "only examiners in charge for screening/assessment and data analysis were blinded"
Incomplete outcome data (attrition bias) All outcomes	Low risk	Judgement comment: attrition was reported on. Numbers of individuals left were all included and outlined within the study flow diagram and results
Selective reporting (reporting bias)	Low risk	Judgement comment: measure proposed was evaluated (FMA‐UE)
Other bias	Low risk	None noted

*Study characteristics*
Methods	Study design: RCT Group: parallel group Number of groups: 2
Participants	Baseline characteristics Experimental: group A (task‐specific motor imagery with mental practice group) age (years): 51.93 ± 5.63 time since stroke (months): 4.73 ± 1.09 sex: 11 men, 4 women stroke severity: Brunnstrom stage 4.13 ± 0.91, Modified Ashworth Scale 7 participants = 1, 8 participants = 1+ Experimental: group B (task‐specific mirror therapy group) age (years): 52.67 ± 7.19 time since stroke (months): 5.10 ± 0.84 sex (male:female): 10 men, 5 women stroke severity: Brunnstrom stage 4.20 ± 0.67, Modified Ashworth Scale 6 participants = 1, 9 participants = 1+ Inclusion criteria: "subjects included in the study were unilateral hemiplegic stroke right or left, subjects with subacute hemiplegia between 2 to 6 months post stroke, ischemic stroke, age above 45 years and below 80 years, both male and female subjects, Brunnstrom stage of motor recovery of 3 to 5, Modified Ashworth Scale score < 2" Exclusion criteria: "subjects were excluded with wrist and/or finger contracture, significant visual, auditory impairment, acute and chronic stroke, subjects with behavioral and attention impairments, global aphasia with cognitive impairments that might interfere with understanding instructions"
Interventions	Experimental: group A (task‐specific motor imagery with mental practice group) frequency: 60 minutes, 3 times/week × 10 weeks plus home exercises on alternate days components: "task‐specific motor imagery with mental practice incorporating physical practice with videotape"; sitting "comfortably in a chair in front of the table containing task‐related materials (videotape displayed on the computer screen, empty paper cup without handle, book, phone, and pen). The affected limb and unaffected limb [were] placed on the table. The subject was first asked to observe the videotape of motor tasks and was instructed to physically practice each task for ten times. Then the subject was asked to mentally practice each task ten times with an interval of five minutes between them. The subject was instructed to indicate the beginning and ending of each task saying 'GO'." Tasks included picking up and putting down a paper cup, turning book pages, holding a pen, picking up a phone, touching top of the head. Home program: completed on days not doing MP: stretching, strengthening upper and lower extremities, balance exercises in standing, functional tasks of upper extremity number of minutes total: 1800 (plus home exercise program) Experimental: group B (task‐specific mirror therapy group) frequency: 60 minutes, 3 times/week × 10 weeks plus home exercises on alternate days components: "positioned on a height adjustable table comfortably with the mirror accommodated between the affected and unaffected limbs. The affected limb was placed behind the mirror and unaffected limb in front of the mirror. The mirror was positioned in front of the patient’s midline, so that the affected limb was fully covered by the mirror and the reflection of the unaffected limb was completely visible. The subject was first demonstrated to perceive the limb differently to how it actually is. Next, subject was instructed to observe the mirror reflection for one to two minutes, trying to visualize the mirror image as the affected limb. Once the subject got engaged with the mirrored limb, they were asked to perform slow, easy to achieve bilateral movements (perceived bilateral movements) while continuing to look at the reflected image." The tasks were the same as for the MP group. Home program: completed on days not doing mirror therapy: same as for the MP group number of minutes total: 1800 (plus home exercise program)
Outcomes	ARAT: measures grasp, pinch, grip, gross movement outcome type: continuous outcome direction: higher is better
Identification	Sponsorshipsource: none noted Country: India Setting: hospital, outpatient Author's name: Sai Kumar N, MPT Institution: Professor & Principal, K.T.G. College of Physiotherapy and K.T.G. Multi Speciality Hospital Address: K.T.G. College of Physiotherapy and K.T.G. Multi Speciality Hospital, Bangalore‐560 091, India
Notes
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "simple random sampling method using closed envelopes"
Allocation concealment (selection bias)	Low risk	Quote: "simple random sampling method using closed envelopes"
Blinding of participants and personnel (performance bias) All outcomes	High risk	Judgement comment: it is unclear if there was blinding of participants and personnel. It is unclear if personnel were blinded or not, as it is unclear if participants just watched a videotape. It appears as though there was study personnel present, and would be able to tell which group was which, as per what participants were doing
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Judgement comment: there was no mention as to whether or not outcome assessors were blinded
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Judgement comment: 30 individuals at the beginning of the study, randomized into 2 groups of 15 participants each. In the results, tables do not give an n number, the study does not mention whether all participants were analyzed in the final results, and the study does not mention whether there were any dropouts
Selective reporting (reporting bias)	Low risk	Judgement comment: it was outlined that the only outcome measure being used in the study was the ARAT. It was also mentioned that this was a limitation of the study, in that it was the only outcome measure used
Other bias	Low risk	None noted

Study	Reason for exclusion
BovendEerdt 2010	Not stroke or mixed diagnoses
Hemmen 2007	Intervention was not imagining motor skills or activities; only 1 isolated muscle movement was imagined
Liu 2004	Not upper extremity or combined upper and lower extremity
Miltner 1999	Article in German; when translated, we found that this was not an RCT

Methods	"Randomly case control study"
Participants	39 people with stroke (20 experimental, 19 control)
Interventions	Motor imagery plus therapy
Outcomes	FMA and modified BI
Notes	Only abstract in English ‐ no interpreter available

Methods	Random allocation
Participants	Unsure; accessed in PEDro ‐ could not access abstract
Interventions	Unsure
Outcomes	Unsure
Notes	Only abstract in English ‐ no interpreter available

Methods	"Randomly divided"
Participants	40 people with stroke
Interventions	FES and motor imagery vs FES
Outcomes	FMA and ARAT, AROM
Notes	Only abstract in English ‐ no interpreter available

Methods	"Random parity on admission"
Participants	95 with stroke
Interventions	Mental imagery, EMG biofeedback, mental imagery + EMG biofeedback
Outcomes	FMA, modified BI
Notes	Only abstract in English ‐ no interpreter available

Study name	Graded motor imagery based on mirror neuron on rehabilitative training for stroke patients: a BOLD‐fMRI study
Methods	Study design: RCT Study grouping: parallel group Number of groups: 2
Participants	15 participants in each group Inclusion criteria: (1) participants signed informed consent; (2) unconscious obstacles, condition is relatively stable, no obvious lack of eyesight; (3) aged 40 to 75 years; (4) no history of cerebrovascular disease; (5) cerebral infarction diagnosis standards, the course in 2 weeks 3 months, right‐handed, left hemiplegia; (6) no metal implants in the body, no MRI testing taboo; (7) National Institutes of Health Stroke Scale score > 4 minutes, paresis test positive, muscle strength level 1 to 3 Exclusion criteria: (1) not diagnosed with cerebral infarction by imaging; (2) acute stage of cerebrovascular disease, unstable vital signs; (3) serious mental illness; (4) with understanding disabilities who cannot meet the test; (5) with serious heart, liver, and kidney dysfunction; (6) with contraindications to MRI examination
Interventions	Intervention characteristics Experimental: routine rehabilitative care + graded motor imagery Control: routine rehabilitative care
Outcomes	FMA‐UE Nine Hole Peg Test Modified BI ‐ activities of daily living Box and Blocks Test JTTHF
Starting date	Ongoing study
Contact information	Author's name: Gu Pengpeng Institution: The Second Affiliated Hospital of Wenzhou Medical University Email: 425575719@qq.com Address: 109 West Xueyuan Road, Wenzhou, China
Notes	ChiCTR IOR 16008137 2016

Study name	Mental imagery combined with physical practice to enhance hand recovery in stroke patients ‐ a randomized controlled trial
Methods	Study design: RCT Study grouping: parallel group Number of groups: 2
Participants	30 participants in each group. Inclusion criteria: "(1) patients diagnosed with stroke (old cases admitted in the stroke recovery ward and new case) in JIPMER. (2) Paralysis limited to one side. (3) Nadir hand function level > Brunnstrom III. (4) Sufficiently preserved comprehension, attention, and intelligence. (5) Patients who have preserved abilities to speak and understand Tamil or Malayalam or English (languages that the instructor is familiar [with]). (6) Age more than 18 years to 65 years" Exclusion criteria: "(1) subjects with symptomatic carotid artery stenosis/with severe cardiac problems/other medications which need immediate intervention. (2) Subjects with persistent language deficit, neglect/inattention. (3) Those with dyspnea, cardiac failure, or other debilitating systemic illness, who cannot perform the physical practices involving the procedure. (4) Patients with significant visual or hearing impairments. (5) Patients who are taking selective serotonin reuptake inhibitors/benzodiazepines and those diagnosed with depression or other major psychiatric disorders or cognitive impairment or dementia"
Interventions	Intervention characteristics Experimental: mental imagery plus routine physiotherapy Control: routine physiotherapy
Outcomes	Hand function
Starting date	Ongoing study
Contact information	Author's name: Unni Krishnan SV Institution: JIPMER ‐ Department of Neurology Email: sunil.narayan@jipmer.edu.in Address: College Of Nursing ‐ JIPMER Dhanvanthari Nagar Gorimedu Pondicherry, 6 Prof Dr Sunil, K Narayan HOD ‐ Department of Neurology JIPMER Pondicherry 605006 India
Notes	CTRI/2016/05/006930 2016

Study name	Effectiveness of graded motor imagery (GMI) on upper limb motor function in patients with subacute stroke
Methods	Study design: RCT Study grouping: parallel group Number of groups: 2
Participants	Sample size: 146 Inclusion criteria Patients between 4 weeks and within 12 months post stroke Patients with first episode of ischemic or hemorrhagic unilateral stroke Patients residing in selected community area registered under pain and palliative society Patients aged between 31 and 70 Patients with subacute stroke exhibiting stable, residual motor deficits in their affected limbs Patients with right or left arm paresis Patient with ability to imitate action with ipsilateral arm Patients who are able to respond to verbal instructions Patients with MIQ‐RS score with ability to do motor imagery Able to sit independently for > 30 minutes Patients who had no previous exposure to mirror therapy Exclusion criteria Patients with cognitive deficits (MMSE < 24) Patients with comorbidities, with apraxia or aphasia Patients with other neurological or orthopedic disorders affecting upper extremity motor function Individuals with any musculoskeletal disorders, visual impairment, systemic disease, and neurological disorder other than stroke Non‐co‐operative patients and suffering from psychological problems
Interventions	Intervention characteristics Graded motor imagery
Outcomes	Upper limb motor function Depression, quality of life, visuospatial neglect, body image
Starting date	22 March 2019
Contact information	Author's name: Prof Fatima D Silva Institution: Nitte Usha Institute of Nursing Sciences, Paneer, Kotekar Beeri Road, Deralakatte, Mangalore 575018, Dakshina Kannada, Karnataka, India Email: ftds_1970@rediffmail.com Address: Paneer, Kotekar Beeri Road, Deralakatte, Mangalore 575018 Dakshina Kannada, Karnataka, India
Notes	CTRI/2019/01/016966

Study name	Effects of adjuvant mental practice using inverse video of the unaffected upper limb in subacute stroke: a randomized controlled trial
Methods	Study design: RCT Study grouping: parallel group Number of groups: 2
Participants	Inclusion criteria: "first‐ever stroke (infarction or hemorrhage diagnosed with brain CT or MRI); hemiplegia with unilateral lesion; subacute status, stroke occurred > 1 months and < 6 months; MMSE ≥ 24 (to be able to understand and follow instructions); age ≥ 18" Exclusion criteria: "severe spasticity, with a Modified Ashworth Scale ≥ 3; excessive pain in the affected upper limb, Visual Analog Scale ≥ 5; enrolled in any experimental rehabilitation studies; poor general condition; not willing to participate in study"
Interventions	Intervention characteristics Experimental: mental practice plus conventional therapy Control: conventional therapy
Outcomes	Fugl‐Meyer upper limb Manual function test Functional Independence Measure
Starting date	Ongoing study
Contact information	Author's name: Hyun Im Moon Institution: Bundang Jesaeng Hospital Address: 20, Seohyeon‐ro 180 beon‐gil, Bundang‐gu, Seoungnam‐si, 13590, Gyeonggi‐do, Republic of Korea
Notes	KCT0002574 2017

Study name	Mental imagery to reduce motor deficits in stroke
Methods	Study design: randomized cross‐over trial Study grouping: parallel group Number of groups: 2
Participants	20 participants Inclusion criteria Minimum motor criteria active wrist extension of no less than 10°, 10° of thumb abduction/extension, and at least 2 additional digits. Must be able to perform at least 3 times in 1 minute Passive range of motion at least 90° of shoulder flexion and abduction, 45° of shoulder external rotation, no less than ‐30° of elbow extension, 45° of forearm supination (from neutral), 45° of forearm pronation (from neutral), wrist extension to neutral, finger extension (all digits) such that no MCP joint has greater than a 30° contracture Participants who are 3 to 12 months post stroke (first‐time clinical CVA) of ischemic or hemorrhagic type at the beginning of the intervention Participants with sensory loss will be eligible Participants must score at least 24 on the MMSE. Memory deficits associated with dementia are the main concern with this inclusion criterion Balance criteria: participants must be able to transfer to and from the toilet independently and safely, stand from a sitting position, and maintain standing balance independently for at least 2 minutes with or without their own upper extremity support Each participant must have physician approval for participation in the therapy to rule out major medical problems Participants must be over the age of 18 years. No upper age limits will be set Participants must have sufficient endurance and stamina determined by clinical judgement to carry out the requirements of the CIT trial Participants with unilateral perceptual neglect will be accepted if they meet the other inclusion criteria Participants on medication will not be excluded. Patients will be excluded if they have received Botox injections to their affected upper extremity less than 3 months before participation, if they have received phenol block injections less than 12 months before participation, if they are receiving intrathecal baclofen, or if they are receiving baclofen or dantrium orally at the time of the study Transportation to and from Emory Center for Rehabilitation Exclusion criteria Score < 24 on the MMSE First stroke less than 3 months or more than 12 months before enrolment Younger than 18 years of age Clinical judgement of excessive frailty or lack of stamina (e.g. cannot attend to instructions, stay awake, engage in functional activities) Serious uncontrolled medical conditions Excessive pain in any joint or more affected extremity that could limit ability to co‐operate with the intervention, as judged by the examining clinician Passive range of motion < 90° of shoulder flexion and abduction, < 45° of shoulder external rotation, > 30° flexion contracture of elbow, < 45° of forearm supination and pronation from neutral, less than neutral wrist extension, and > 30° flexion contracture of any MCP joint Unable to stand independently for 2 minutes, transfer independently to and from the toilet, or perform sit‐to‐stand Have received injections of antispasticity drugs into upper extremity musculature within past 3 months, or wish to have drugs injected in the foreseeable future Receiving any antispasticity drugs orally at the time of expected participation Received phenol injections less than 12 months before participation Damage to the parietal area of the brain
Interventions	Experimental: mental imagery and CIT mental imagery and constraint‐induced therapy: listening to an audiotape to enhance mental imagery + participation in CIT Active comparator: mental Imagery only: listening to an audiotape to enhance mental imagery of functional activities of the affected upper extremity
Outcomes	Primary outcome measures Wolf Motor Function Test (time frame: pretreatment, post treatment, and at 3‐month follow‐up) Fugl‐Meyer Motor Assessment Test (time frame: pretreatment, post treatment, and at 3‐month follow‐up) Movement Imagery Questionnaire (time frame: pretreatment, post treatment, and at 3‐month follow‐up) Vividness of Movement Imagery Questionnaire (time frame: pretreatment, post treatment, and at 3‐month follow‐up) Center for Epidemiologic Studies Depression Scale (time frame: pretreatment, post treatment, and at 3‐month follow‐up) Stroke Impact Scale (time frame: pretreatment, post treatment, and at 3‐month follow‐up) Secondary outcome measure: Sirigu's break test (time frame: pretreatment, post treatment, and at 3‐month follow‐up)
Starting date	States study completion in August 2011
Contact information	Andrew Butler, PhD, Prinicipal investigator, Emory University
Notes	NCT00379392

Study name	Mental practice in chronic, stroke‐induced hemiparesis
Methods	Study design: RCT Study grouping: parallel group Number of groups: 2 Triple‐blinded
Participants	100 participants Inclusion criteria FMA score > 27, which is indicative of minimal to moderate arm impairment stroke experienced > 6 months before study enrolment minimal cognitive impairment, score > 25 on the Folstein MMSE age > 21 years and < 80 years have experienced 1 clinical stroke as verified by a physician discharged from all forms of physical rehabilitation targeting the affected arm Exclusion criteria < 21 years old excessive pain in the affected hand, arm, or shoulder, as measured by a score > 5 on a 10‐point visual analog scale excessive spasticity in the affected biceps, triceps, wrist, or fingers, defined as a score ≥ 2 on the Modified Ashworth Spasticity Scale currently participating in any experimental rehabilitation or drug studies targeting the upper extremity mirrors movements (i.e. involuntary movements by the unaffected hand during attempts at unilateral movement by the stroke‐affected hand during attempts at unilateral movement by the stroke‐affected hand) history of parietal stroke (because some data suggest that ability to estimate manual motor performance through mental imagery is disturbed after parietal lobe damage) affected arm joint restriction that in the opinion of the investigator would hinder study participation
Interventions	Experimental: mental practice group: patients are administered rehabilitative therapy targeting their affected arms 3 days/week during a 10‐week period. Directly after the therapy session, they participate in targeted mental practice session in which they cognitively rehearse the movement that they just physically practiced Active comparator: active control group: individuals assigned to this group are administrated rehabilitative therapy targeting their affected arms in half of 4 increments, occurring 3 days/week for 10 weeks. They also are administered an intervention in which they listen to a relaxation tape and/or tapes in which they receive instructions on exercises and information on stroke care
Outcomes	Action Research Arm Test
Starting date	States study completion date December 2016
Contact information	Stephen Page, PhD, MS, Associate Professor School of Health and Rehabilitation Sciences Division of Occupational Therapy, Ohio State University
Notes	NCT01651533

Study name	Mental practice in post‐stroke subjects
Methods	Study design: RCT Study grouping: parallel group Number of groups: 3
Participants	30 participants Inclusion criteria: "clinical diagnosis of stroke, ischemic or hemorrhagic, for more than 6 months, age above 18 years, unilateral involvement and able to hold objects" Exclusion criteria: "painful conditions that affect the ability to perform the proposed exercises, spasticity greater than 3 by the Ashworth Scale, and cognitive deficits that will be evaluated by the Mini Mental State Examination (MMSE)"
Interventions	Intervention characteristics Mental practice before physical practice Physical practice before mental practice Physical practice only
Outcomes	Action Research Arm Test: 8 weeks and 3 months follow‐up Fugl‐Meyer Assessment ‐ Upper Extremity Component: 8 weeks and 3 months follow‐up Surface electromyography, short radial extensor of the carpus, and superficial flexor of the fingers at 8 weeks and 3 months follow‐up Functional Independence Measure: 8 weeks and 3 months follow‐up Box and Block test: 8 weeks and 3 months follow‐up Movement Imagery Questionnaire ‐ Revised second version (MIQ‐RS): 8 weeks Kinesthetic and Visual Imagery Questionnaire (KVIQ‐10): 8 weeks Theory of Mind Task Battery (ToM): 8 weeks Modified Ashworth Scale (MAS): 8 weeks and 3 months follow‐up
Starting date	Ongoing study
Contact information	Author's name: Roberta O Cacho Institution: Faculty of Health Science ‐ Facisa/UFRN Address: Santa Cruz, Rio Grande do Norte, Brazil, 59200000
Notes	www.clinicaltrials.gov/ct2/show/NCT03251209

PERMALINK

Mental practice for treating upper extremity deficits in individuals with hemiparesis after stroke

Ruth E Barclay

Ted J Stevenson

William Poluha

Brenda Semenko

Julie Schubert

Abstract

Background

Objectives

Search methods

Selection criteria

Data collection and analysis

Main results

Authors' conclusions

Plain language summary

Summary of findings

Background

Description of the condition

Description of the intervention

How the intervention might work

Why it is important to do this review

Objectives

Methods

Criteria for considering studies for this review

Types of studies

Types of participants

Types of interventions

Types of outcome measures

Primary outcomes

Secondary outcomes

Search methods for identification of studies

Electronic searches

Searching other resources

Data collection and analysis

Selection of studies

Data extraction and management

Assessment of risk of bias in included studies

Measures of treatment effect

Unit of analysis issues

Dealing with missing data

Assessment of heterogeneity

Assessment of reporting biases

Data synthesis

Subgroup analysis and investigation of heterogeneity

Sensitivity analysis

GRADE and "Summary of findings"

Results

Description of studies

Results of the search

1.

Included studies

Excluded studies

Ongoing studies

Studies awaiting classification

Risk of bias in included studies

2.

3.

Allocation

Blinding

Incomplete outcome data

Selective reporting

Other potential sources of bias

1.1. Analysis.

1.2. Analysis.

1.3. Analysis.

4.

5.

6.

Effects of interventions

Summary of findings 1. Mental practice in addition to other treatment compared to other treatment (± placebo) for treating upper extremity deficits in individuals with hemiparesis after stroke.

Summary of findings 2. Mental practice compared to conventional therapy for treating upper extremity deficits in individuals with hemiparesis after stroke.

Comparisons

2.1. Analysis.

2.2. Analysis.

3.1. Analysis.

3.2. Analysis.

4.1. Analysis.

4.2. Analysis.

Results of meta‐analyses

Study name	Effectiveness of first‐person and third‐person motor imagery in relearning daily hand tasks for people with chronic stroke: a randomized controlled trial
Methods	Study design: RCT Study grouping: parallel group Number of groups: 3
Participants	20 participants in each group. Inclusion criteria: "(1) age between 18 and 80 years. (2) A diagnosis of hemiplegia due to stroke (infarct or hemorrhage). (3) Stroke experienced more than 3 months prior to enrolment (Summers et al, 2007). (4) Able to complete Advanced Hand Activities 1, 2, and 3 of Motor Assessment Scale (Carr et al, 1985). (5) Score of more than 24 points on the Mini‐Mental State Exam (dictates no cognitive deficits) (Folstein et al, 1975). (6) Able to give voluntary consent to participate in the study" Exclusion criteria: "(1) visual and perceptual problems including hemianopia and unilateral neglect. (2) Excessive pain in the affected arm as measured by score of more than 4 on a 10‐point Visual Analog Scale (Scott & Huskisson, 1976). (3) Spasticity of more than 3 on Modified Ashworth Scale (Bohannon & Smith, 1987). (4) Pre‐existing musculoskeletal, neurological (apart from stroke), or other conditions that may affect upper limb function"
Interventions	Intervention characteristics First person imagery Third person imagery Conventional occupational therapy
Outcomes	JTTHF Motor Activity Log: amount of use Motor Activity Log: quality of movement Vividness of Movement Imagery Questionnaire Fugl‐Meyer Motor Assessment Nine Hole Peg Test Inhand Manipulation Assessment Lawton Instrumental Activities of Daily Living Canadian Occupational Performance Measure Stroke Specific Quality of Life
Starting date	Ongoing study 2017‐09‐25
Contact information	Author's name: Dr Padma S Gunaratne Institution: Department of Neurology, The National Hospital of Sri Lanka Email: pagunara@gmail.com Address: Regent Street, Colombo 10, Sri Lanka
Notes	SLCTR/2017/031 2017