A computer-game-based rehabilitation platform for individuals with fine and gross motor upper extremity deficits post-stroke (CARE fOR U) – Protocol for a randomized controlled trial

Abstract

Background & purpose

Activity-based neuroplasticity and re-organization leads to motor learning via replicating real-life movements. Increased repetition of such movements has growing evidence over last few decades. In particular, computer-game-based rehabilitation is found to be effective, feasible and acceptable for post-stroke upper limb deficits. Our study aims to evaluate the feasibility and effectiveness of 12 weeks of computer-game-based rehabilitation platform (GRP) on fine and gross motor skills post-stroke in India.

Methods

Through this trial we will study the effect of adjunctive in-hospital GRP (using a motion-sensing airmouse with off-the-shelf computer games) in 80 persons with subacute stroke, for reduction of post-stroke upper limb deficits in a single-centre prospective Randomized Open, Blinded End- point trial when compared to conventional therapy alone.

Results

We intend to evaluate between-group differences using Wolf Motor Function test, Stroke Specific Quality of Life, and GRP assessment tool. Feasibility will be assessed via recruitment rates, adherence to intervention periods, drop-out rate and qualitative findings of patient experience with the intervention.

Conclusion

The CARE FOR U trial is designed to test the feasibility and effectiveness of a computer-game based rehabilitation platform in treating upper limb deficits after stroke. In case of positive findings GRP can be widely applicable for stroke populations needing intensive and regular therapy with supervision.

Introduction & rationale

Over the last few decades, there has been an increase in stroke burden globally. Global burden of disease (GBD) 2016 reports 80.1 million prevalent cases of stroke globally. and 116.4 million disability adjusted life years. ¹ Asia accounts for almost two-thirds of the world’s stroke incidences. ²

A significant proportion of individuals with stroke have long term residual disability ³ and three-fourths of Indian stroke survivors are left with moderate to severe disabilities. ⁴ Loss of arm function with stroke can impact a person’s ability to participate in home life, work and can reduce quality of life. Upper limb (UL) impairments are seen in 80% of stroke survivors and only 5–20% have complete functional recovery.^5–7 with residual spasticity in 46% of cases ⁵ and only 20.7% of stroke survivors returning to work by 2 years post-stroke (with half having changed their job). ⁸ Early and targeted rehabilitation for UL deficits post-stroke is therefore crucial. ⁹

Current evidence-based approaches used in rehabilitation of UL function after stroke emphasize that intensity, volume of training and task-specificity are pivotal. ⁵ One such approach, the constraint-induced movement therapy, established strongly to be effective in rehabilitation of UL deficits, requires constraint of the unaffected limb and performance of repetitive functional training of upper extremity. This constraint can be applied for up to 60–90% of waking hours^10,11 Questions arise on the ability of patients to adhere to such therapy regimes which are tedious and may lead to poor adherence in those with low motivation ¹² which is essential for improved compliance and adherence to rehabilitation regimes for individuals with UL deficits.^13–15

Activity-based neuroplasticity and re-organization leading to motor learning by replicating real-life movements and repetition has growing evidence over last few decades.^16,17 Virtual reality has been proven to improve UL function and ADLs when used as an adjunct to usual care (probably by increasing the duration of overall therapy). ¹⁸ Such a mode of therapy has frequently been applied via the use of commercial gaming consoles like the Sony Playstation, Nintendo Wii and Microsoft Kinect. However VR based therapy has its cons: lack of computer skills of therapists, support infrastructure, initial investment, inadequate communication infrastructure in case of telerehabilitation and questionable long lasting effects.^18,19 Studies also provide an increasing evidence for use of gaming technology in stroke rehabilitation of UE.^20,21 Computer based systems may allow less dependence on rehabilitation personnel, improved standardization of rehabilitation protocols, increased intensity and frequency of activities and creative treatment delivery. ²² A structured game-based rehabilitation protocol could effectively improve post-stroke upper limb function^15,23,24 and is reported to be feasible, well-accepted by patient population with no adverse effects. ²¹

Our study will therefore investigate the feasibility and effectiveness of a simplified in-hospital upper-limb rehabilitation program using off-the-shelf computer-games and a miniature wireless motion detecting mouse (Airmouse™) mounted on various objects to replicate activities of daily living (ADLs) for those with subacute stroke. We hypothesize that adjunctive Computer game-based rehabilitation platform for post-stroke upper extremity fine and gross motor deficits, will be feasible and effective in treating UL deficits due to stroke when compared to conventional therapy alone in the subacute phase

Methods

CARE FOR U

A computer-game-based rehabilitation platform (GRP) previously developed for assessment and rehabilitation of UL functions of individuals with stroke and cerebral palsy will be used in this trial for those with stroke.^25,26 This GRP includes three main components:

1. A miniature wireless inertia-based motion-detecting mouse which translates movements performed to the motion of a computer cursor: Airmouse™ (Figure 1). When mounted on objects, the mouse is used to control the motion of the cursor/game paddle. The Airmouse™ can be mounted on a varied range of objects of differing physical properties (weight, texture, shape, size) that create diverse functional demands allowing for practice of manual dexterity, precision and power grasps.

2. A software-based assessment game with data logging and analysis methods. Electronic records are generated and stored digitally, which can be read as meaningful graphs and tables for reviewing patient progress. (Figure 2)

3. Off-the-shelf computer games for therapy sessions.^27,28

Figure 1.

AirmouseTM (In comparison to a standard optical computer mouse).

Figure 2.

AirmouseTM (In comparison to a standard optical computer mouse).

GRP follows multiple principles of motor learning namely, task and goal oriented therapy and multisensory stimulation. We believe that this form of therapy will be useful in low and middle income countries (LMICs) like India where adherence to rehabilitation post-stroke is poor ²⁹

Design

A Phase-II randomized controlled trial with PROBE (Prospective Randomized Open, Blinded End- point) design. ³⁰

Inclusion criteria

1. Between 18-70 years of age

2. Individuals with first ever stroke

3. Time of stroke onset between 1-3 months

4. Level of spasticity in upper extremity between grades 1 and 1+ based on Modified Ashworth scale ³¹

5. Brunnstrom hand recovery stages of 4, 5, 6, shoulder and elbow recovery stages of 3,4, 5 ³² (Supplementary File 1)

Exclusion criteria

1. Individuals scoring <20 on Mini mental state examination.^33,34

2. Individuals with visual, auditory, perceptual issues

3. Individuals having premorbid neurological deficits

4. Individuals having a premorbid modified Rankin scale >3

5. Those unwilling/unable to participate

Randomization and recruitment

Eligible participants will be screened by the trial coordinator/Intervention therapist. Informed consent will be received from all recruited participants according to the International Conference on Harmonization Good Clinical Practices guidelines (ICH/GCP)^{. 35} and ethical guidelines for biomedical research on human participants published by the Indian Council of Medical Research, New Delhi. ³⁶ Following baseline assessments participants will then be randomized either into the intervention or control arm in a 1:1 ratio using protected, web-based Random Allocation software.

Screening and baseline assessments

All participants attending stroke and Neurology outpatient department will be screened by a trial coordinator/Intervention therapist based on the above mentioned inclusion criteria. Those fulfilling the inclusion criteria will be consented and baseline scores (Immediately after randomization) of the following scales will be recorded: Wolf motor function test, Stroke specific quality of life scale, Brunnstrom staging of hand and shoulder-elbow along with the GRP assessment game outcomes (see below for explanation) (Figure 3).

Figure 3.

Process Flow.

Outcome measures

Final outcomes will be measured at 12th week post-randomization and follow up at 16th week post randomization

Primary outcomes

1. Wolf motor function test (WMFT):^37,38 A 17-item scale quantifying upper extremity movement ability through timed single- or multiple-joint motions and functional tasks. Each task performed as quickly as possible and is graded based on movement quality on observation on an ordinal scale of 0 to 5, where 0 indicates no performance and 5 indicates performance with normal movement. The final WMFT scores will be the total time taken for the 15 tasks and the summed movement quality grades of the 17 tasks.

2. Stroke specific quality of life (SS-QOL):^39,40 In this patient centered questionnaire for health-related quality of life specific to post-stroke individuals, individuals respond to 49 questions under 12 domains with reference to the past week. Responses are rated on a 5-point likert scale with a higher score indicating better function. Like above same applies here

Secondary outcome

GRP assessment tool is a computer-based assessment tool to quantify object manipulation skills. The software displays a target object appearing at random locations at the top of the display and moving to the bottom within two seconds to disappear. One ‘game event’ is defined as the time between the appearance of a target and its disappearance. The patient is required to move 5 chosen objects (Figure 4 with task description,) mounted with the Airmouse™ in order to move the game paddle and catch the moving target before it hits the bottom of the screen. These 5 objects are chosen to assess various functional and anatomical demands. The upper limb movements of every ‘game event’ are displayed as graphical outputs on the screen. These can be quantitatively measured as response time, movement time, movement error, movement variance and total success rate. (Figures 2 and 4)

Figure 4.

Five Assessment Tasks.

Intervention arm

Participants randomized to the intervention arm will receive conventional therapy and in-hospital therapy for 12 weeks, 3 sessions/week by the intervention therapist as face-to-face sessions. Intervention will include:

1. Conventional therapy as followed in the hospital (see below)

2. GRP intervention: The patient will be required to play Off-the-shelf computer games (Aquaball and Brave-Piglet, Figure 5) while manipulating objects of various physical properties (like size, weight, texture, and shape and surface properties) with Airmouse™ mounted on them. This allows practice of different modes of manipulation and functional tasks by taking advantage of the ergonomic properties of chosen objects (Figure 6). The initial exercise protocol for each participant will be tailor-made based on patient’s goals, level of impairment and functional status. Therapy will be progressed by increasing the difficulty level of the game being played according to individual patient needs (Increasing speed of the task, increasing number of distractors, modifying the size/weight/shape/texture of the object to be manipulated). GRP also allows for incorporation of movement precision. Numerous choices of the objects to be manipulated and the choice of the off-the-shelf game to be played provide the scope to tailor-make exercise regimes for each patient, thus increasing its therapeutic value. ²⁵ Dosage of one session lasting for 45–60 minutes, thrice a week, for a period of 12 weeks will be followed.

Figure 5.

Off-the-shelf computer games.

Figure 6.

Object choices for exercise regime.

The following are of the two commercial computer games which will be used in GRP intervention (Figure 5)

1. Aquaball: This is a brick buster game with slow to moderate speed and low to moderate movement precision requirement. It present with few distractors and provides simple to complex 2D backgrounds. ²⁵

2. Brave Piglet: This is a vertical single-axis shooting game with moderate speed levels, low to moderate level of distractors and multiple targets. ²⁵

Control group

Participants randomized to the control arm will receive conventional therapy followed by the Neuro-rehabilitation department in Christian Medical College and Hospital, Ludhiana, for 12 weeks, 3 sessions/week. The control group therapy includes a combination of various neurological approaches (including but not limited to Task oriented approach, Proprioceptive neuromuscular facilitation techniques etc.) of stroke rehabilitation. The current format of rehabilitation sessions in this center is custom-made for every patient according to their needs and lasts for 30–45 minutes/session.

Treatment provided to both groups will be recorded in case record forms, details of therapy type, intensity, frequency and total duration will be recorded.

Feasibility

A qualitative assessment of intervention acceptability will be conducted by in-person follow-up at the end of the 12 week program. The following open ended questions will be administered to participants randomized into the intervention group:

1. How do you think this rehabilitation has helped you?

2. Has this form of rehabilitation helped you to adhere to the protocol regime better?

3. Did you enjoy playing the games?

4. Are you able to carry out your daily activities in an improved and easier way than before?

5. Will you refer this rehabilitation method to others in the future?

Adverse events

Any adverse event will be recorded through a checklist at every follow-up and reported to the Institutional ethics committee. In case the patient complains of pain or fatigue at the end of intervention sessions, they will be assessed for pain on Visual Analogue Scale and physical fatigue on Borg’s scale of perceived exertion. The other expected adverse events may be irritability if unable to complete the game and headache due to prolonged computer-screen exposure. Serious adverse events (those requiring medical attention or re-hospitalization) will be reported to the institutional ethic committee. Safety of the study will be established by noting and comparing adverse events in all recruited participants.

Prevention of risks to internal validity

To prevent diffusion/contamination of intervention to the control group, we will allow intervention and control therapy sessions to be held at varying times or in varying locations at the recruiting hospital. Blinding will be maintained by appointing a dedicated blinded assessor for follow-up of all recruited participants and by performing follow-up assessments away from the treatment location. All participants will be advised during recruitment and treatment sessions to avoid talking about their treatment sessions with the blinded assessor. Intervention fidelity will be maintained by regular compliance checks by the Principal Investigator (PI), maintenance of activity logs performed by the intervention therapist and ensuring similar conventional therapy sessions for both intervention and control groups. This will be periodically cross checked by the PI by monitoring the hospital patient records

Ethics

Ethical approval has been obtained from the institutional Ethics Committee at Christian Medical College and hospital Ludhiana. Study protocol (version-1), Case report forms 1 and 2 (version-1) and Informed consent forms in English, Hindi and Punjabi (Version-1.2), have been approved for use in the study.

Sample size and power

The delayed intervention arm in EXCITE study showed a 26% change in their primary outcome (WMFT). ¹⁰ Thus, with GRP intervention for our study, a sample size of N = 40 was calculated for each group, expecting the primary outcome clinical difference of 25% at 90% power, type 1 error rate of 5% and 20% drop-out rate.

Statistical methods and results

We intend to evaluate the differences between the intervention and control groups based on Wolf Motor Function Test (Activity limitation) at 12th and 16th weeks post-randomization; Stroke Specific Quality of Life (Participation Restriction) and GRP assessment tool at the 12th and 16th week follow up post-randomization. Scores of participants completing a minimum of 8 therapy sessions over a period of 5 weeks (Intervention or control) will be considered for analysis. Modified intention-to-treat principle will be used for statistical analysis. Missing values on outcomes will be imputed using last observation carried forward method. Statistical analysis will be performed with SPSS version 21 (IBM Corp., Armonk, NY). P value <0.05 will be considered as statistically significant. Frequencies (with percentages) for binary and categorical variables and means (and standard deviations), or medians (lower and upper quartiles) for continuous variables will be presented. Baseline characteristics of participants will be summarized for both groups. There will be no tests of statistical significance or confidence intervals for differences between groups on any baseline variables. The primary outcomes will be analyzed using a mixed effect regression model to account for repeated measures over time. The mixed effect model will include change (from baseline to follow-up) in primary outcome as the response variable with randomized group, time and an interaction between time and randomized group as a fixed effect to allow estimation of the treatment effect at both follow ups. To check the possible differences of treatment effect at each time point, interaction term between time and randomized group will be included. The model will be adjusted for age, gender and baseline outcome data.

Feasibility of the study will be evaluated on the recruitment rate (a minimum of 70%), completion of/adherence to intervention periods (a minimum of 8 sessions over a period of 5 weeks), drop-out rate (<20%) and questionnaire (see above) based-qualitative findings of patient experience with the intervention. Patient responses (recorded in local languages as per patient preference) will be translated and transcribed by certified personnel. Data will then be categorized and presented as a descriptive report to understand facilitators and barriers of the study intervention.

The intervention will be considered safe if no difference in adverse events related to upper extremity training between the two groups.

Discussion

Task-oriented therapy approaches which are challenging, intense and repetitive have the potential to induce neuroplasticity for recovery.^21,41 The feasibility, motivation and simplicity of Computer-game-based rehabilitation can be creatively and positively applied in rehabilitating post-stroke individuals. The advantage of using GRP is the extent to which therapy can be customized for every patient by using varied objects of daily living according to the patient’s need. The small-size of Airmouse™ makes it feasible to be mounted over objects of any shape or size allowing multiple variations in therapy regime. This variability in tasks during motor learning improves subsequent motor performance.^42,43 Goal-directed rehabilitation protocols that include visual feedback (graphical outputs in GRP, off-the-shelf gaming visuals), sensory feedback (through objects being held that have varied ergonomic properties) and use of cognitive functions makes GRP advantageous.^44,45 GRP also allows for inclusion of task dynamics and precision movements during therapy ²⁷ which have been proven to stimulate ipsi- and contra-lesional cortices promoting neuro-adaptation in recovery. ⁴⁶

The CARE-FOR-U trial is designed to test the feasibility and effectiveness of a computer-game based rehabilitation platform in treating upper limb deficits after stroke. In case of positive findings GRP can adapted for stroke populations needing intensive and regular therapy. With the added possibility of adapting this platform to a home-environment, tele-rehabilitation with remote supervision by the Physiotherapist would make this form of rehabilitation widely applicable; especially for those with poor access to rehabilitation facilities or difficulty in travelling on a regular basis to a rehabilitation set-up.

ORCID iDs

Dorcas BC Gandhi https://orcid.org/0000-0002-9616-7870

Jeyaraj D Pandian https://orcid.org/0000-0003-0028-1968

Mahesh P Kate https://orcid.org/0000-0003-4794-2882