Mirror Therapy for Lower-Extremity Hemiparesis: A Knowledge Translation Study Using an Educational Module to Change Physiotherapists’ Perceptions

Abstract

Purpose: Despite growing evidence that mirror therapy (MT) is effective for improving lower-extremity (LE) function in patients with stroke, it is not commonly used by physiotherapists. The purpose of this study was to determine whether change would occur in physiotherapists’ knowledge of, confidence in performing, and willingness to use MT for LE stroke rehabilitation after participating in a 1-hour educational module. Method: A convenience sample of physiotherapists working in neurorehabilitation was recruited for a single-group quasi-experimental pre–post study. Participants attended a 1-hour educational module on MT. Therapists’ perceptions of the use of MT were assessed by questionnaire before and after they participated in the module. A follow-up telephone survey was conducted after 3 months to determine how many participants had actually used MT in their practice. Results: Nine physiotherapists participated in this study. Statistically significant increases were found in their perceived knowledge of, confidence in, and willingness to use MT. At the 3-month follow-up, three participants had used MT with patients with LE hemiparesis. Conclusions: Therapists’ knowledge of and attitudes toward MT for LE stroke rehabilitation changed favourably after participating in a 1-hour educational module. The module also led some therapists to make a change in practice at 3 months.

Stroke is a leading cause of long-term disability in Canada.¹ Approximately 62,000 strokes occur each year, and an estimated 405,000 Canadians live with its long-term effects,² which can include motor, sensory, and cognitive deficits. A common impairment after stroke is reduced walking ability, and although most individuals recover some ability to walk, as many as 36% are unable to recover independent walking function even after rehabilitation.³ Because independent walking after stroke is a determinant of quality of life and a commonly cited goal of individuals with stroke, more effective treatments are needed to improve the rate of walking recovery.^4–7

Because the prevalence and consequences of stroke are so well documented, it is not surprising that research to guide stroke rehabilitation has proliferated. The latest edition of the Evidence-Based Review of Stroke Rehabilitation (EBRSR),⁸ a Canadian initiative that provides an up-to-date review of the latest evidence in stroke rehabilitation research, includes more than 4,500 studies; it is thus a testament to the sheer volume of literature available on the subject. However, despite advances in stroke rehabilitation research, clinical implementation of such advances often lags behind: the underuse of evidence-based rehabilitation strategies and variation in clinical practice have been reported in Canada, the United Kingdom, and Australia.⁹ Although adhering to evidence-based practice (EBP) is associated with improved functional recovery, patient satisfaction, and appropriate allocation of health care resources,^10,11 this discrepancy between research and clinical practice can create a significant gap in the quality of care being provided to people with stroke.

One example of the underuse of evidence-based rehabilitation strategies is mirror therapy (MT) as a treatment for lower-extremity (LE) hemiparesis. MT has been proposed as a simple and low-cost adjunct therapy for improving LE motor function after stroke.¹² It was initially developed in the mid-1990s as a treatment for phantom limb pain and has been used in stroke rehabilitation since 1999.^13,14 MT for stroke rehabilitation involves placing a mirror along the mid-sagittal plane and then instructing the subject to move the non-paretic limb while viewing the mirror image; this movement creates the illusion that the affected limb is moving normally.¹⁵ Although research on MT for stroke rehabilitation has traditionally focused on restoring upper-extremity function, a growing body of evidence has demonstrated the benefits of its use in LE rehabilitation.¹⁵ Since 2015, three systematic reviews have concluded that MT can have a significant effect on lower-limb motor recovery, gait speed, balance, and ankle range of motion.^15–17

Despite such strong evidence that MT is an effective treatment for LE hemiparesis, few resources are available to help clinicians understand and implement it in their practice. Moreover, the EBRSR’s Stroke Rehabilitation Clinician’s Handbook does not yet include MT for LE rehabilitation,¹⁸ and no evidence exists to suggest that clinicians in Canada are using MT in rehabilitation for LE hemiparesis. This situation highlights the challenge of integrating novel evidence-based interventions from the research setting into clinical practice, a process that can take up to 17 years.¹⁹

Research into the facilitators of and barriers to incorporating EBP into physiotherapy has pointed to the importance of knowledge translation (KT), the process of disseminating research findings to clinicians, in closing the gap between research and practice.²⁰ Several studies on implementing EBP in physiotherapy have found that lack of confidence, time, and critical appraisal skills are three of the most notable barriers.²¹ Converting research findings into a more personal, accessible, and comprehensible format using KT strategies can optimize physiotherapists’ uptake of evidence into practice.²²

The purpose of this research project was thus to determine whether a change in physiotherapists’ knowledge and attitudes would occur after they participated in a clinically relevant KT intervention designed to bridge the research-to-practice gap in the area of MT for the rehabilitation of LE hemiparesis. The first objective of this study was to determine whether physiotherapists’ perceived knowledge of, confidence in performing, and willingness to use MT for LE hemiparesis would increase after they participated in a 1-hour educational module. The second objective of this study was to determine whether therapists implemented MT for LE rehabilitation into their practice afterward.

Methods

This study used a single-group quasi-experimental pre–post design to determine the change in physiotherapists’ knowledge of and attitudes toward using MT for LE hemiparesis after taking part in an educational module. It also included a 3-month follow-up telephone interview to monitor any actual change in their practice. Ethical approval for the study was obtained from the University of British Columbia’s Behavioural Research Ethics Board in December 2018.

Participants

Eligible participants were registered physiotherapists providing stroke rehabilitation in a clinical setting, whether in an inpatient facility or an outpatient clinic. They were required to be fluent in English so that they could complete the pre- and post-questionnaires and fully participate in the module. Physiotherapists who had received training in MT for LE stroke rehabilitation in the previous 6 months were excluded from the study.

We emailed the physiotherapy manager at three local neurorehabilitation facilities to distribute an invitation to take part in a student-led study to learn about the use of MT for the rehabilitation of LE hemiparesis in people with stroke. We confirmed eligibility and obtained informed consent from participants who responded to our email.

Intervention development

The KT intervention was a 1-hour learner-centred module, delivered as an in-service workshop during participants’ work hours. We developed the module in the context of Graham and colleagues’ Knowledge to Action (KTA) process model,²³ which has been used by the World Health Organization as a framework for facilitating KT.²⁴ The KTA framework contains two key components: knowledge creation, in which thorough research is conducted and ideas are established and refined for clinicians, and knowledge action, which describes the various activities required for knowledge applications to produce real change.²³ We carried out a literature review but were unable to identify any clinical practice guidelines or other knowledge tools for MT for LE hemiparesis; as a result, part of module development involved creating an informal synthesis of clinical trials and systematic reviews to create user-friendly guidelines for therapists to use when applying the technique.

To address the knowledge action component of the KTA cycle, we developed practice guidelines and tailored them to the local contexts of inpatient (acute or subacute) and outpatient (subacute or chronic) stroke rehabilitation; in addition, we considered barriers such as a lack of time and access to equipment in the module. We then developed the KT intervention in the form of a 1-hour in-person educational module that integrated all these considerations.

We delivered the educational module using the BOPPPS model of lesson planning, which incorporates a Bridge, learning Objectives, a Pre-test, a Participatory learning session, a Post-assessment, and a Summary. This model is beneficial because it places importance on gaining participants’ interest, making the purpose of the educational session well known, assessing participants’ level of understanding, engaging them in an active learning environment in which they are able to practise hands-on skills, and then summarizing the main points to reinforce what has been learned.²⁵ A schedule of the session, including a timeline with section breakdowns, is shown in Appendix 1. We conducted a trial run with four therapists who were not participating in the study and incorporated their feedback to enhance the content of the module and increase participant engagement.

Outcomes

We designed our study to have two outcome measures. The first was a nine-item questionnaire to measure the therapists’ perceived knowledge of, confidence in delivering, and willingness to use MT. The therapists responded to three statements corresponding to each of these three domains, worded in the first person and scored on a 5-point Likert-type scale ranging from strongly disagree to strongly agree. The questionnaire’s development was informed by a similar implementation study that assessed the effectiveness of an educational module on health care practitioners’ knowledge, role legitimacy, and motivation with regard to treating people with diabetes mellitus.²⁶ Two therapists with research training reviewed the initial questionnaire, which we then pilot tested in the trial run of the educational module to ensure that its wording was clear. The final version of the questionnaire used for the study can be found in Appendix 2.

Our secondary outcome measure was whether the therapists had actually implemented MT 3 months after the module was delivered. To measure this, we contacted participants by telephone and asked whether they had used MT to treat LE hemiparesis since participating in the module. They were also asked an open-ended question about why they had or had not done so.

Procedures

The educational module was delivered on two occasions at separate facilities in Vancouver, in January and February 2019; each participant attended one of them. Participants completed a paper copy of the questionnaire immediately before and after participating in the module and were reminded that they would be contacted by telephone in 3 months to assess their use of MT. DRL conducted the telephone interviews, which lasted approximately 5 minutes; use of MT and comments were recorded in notes on a follow-up interview form. After each session, we provided the hosting facility with a frameless free-standing mirror specifically designed for this study.

Statistical analysis

The participants’ socio-demographic characteristics were grouped into categories and described using frequencies and percentages. The responses to the three 5-point Likert-scale questions for each domain (knowledge, confidence, and willingness) were converted to scores ranging from 0 for strongly disagree to 4 for strongly agree, and the scores were then combined to provide a total domain score out of 12. Non-normally distributed data were reported using median and first- and third-quartile values. The differences between the participants’ pre- and post-intervention domain totals were evaluated using a non-parametric-related Wilcoxon signed-rank test, setting statistical significance at p < 0.05. Data were analyzed using IBM SPSS Statistics (Version 22.0, IBM Corporation, Armonk, NY).

The participants’ responses during the follow-up telephone interview regarding their use of MT were categorized as either having or not having used MT in treating LE hemiparesis (at least once). Given the brevity of the conversation and the limited number of participants, we did not conduct a qualitative analysis for the one open-ended follow-up question. Instead, comments as to why the therapists had not used MT in the intervening 3 months as well as comments from those who had used MT were arranged into basic categories for descriptive purposes.

Results

Nine participants were recruited in total: four from private practice neurorehabilitation and five from public inpatient facilities. Their socio-demographic data are shown in Table 1. The majority were female and relatively early in their physiotherapy careers.

Table 1.

Participants’ Sociodemographic Data (N = 9)

Characteristic	No. (%)
Sex
Male	1(11.1)
Female	8 (88.9)
Age, y
20–29	4 (44.4)
30–39	5 (55.6)
Physiotherapy experience, y
≤ 5	4 (44.4)
6–10	5 (55.6)
Workplace setting
Private neurorehabilitation (subacute or chronic)	4 (44.4)
Public inpatient rehabilitation (acute or subacute)	5 (55.6)

All participants completed the pre- and post-intervention questionnaires. Figure 1 illustrates the changes in the nine participants’ scores for each of the three domains from before to after participating in the educational module. Statistically significant changes in knowledge (z = −2.68, p = 0.007), confidence (z = −2.68, p = 0.007), and willingness (z = −2.23, p = 0.026) were found. The pre- and post-intervention median (first–third quartile) scores were 5 (3.5–6.0) and 9 (9.0–10.0) points for knowledge, 5 (4.5–8.5) and 10 (9.0–12.0) points for confidence, and 9 (7.0–10.0) and 9 (9.0–11.5) points for willingness, respectively.

Figure 1.

Changes in scores from pre-test to post-test for the nine participants: (a) knowledge, (b) confidence, and (c) willingness.

The 3-month follow-up data were collected from all participants. Three of the nine (33%) reported that they had used MT since participating in the educational module, and six (66%) reported that they had not. The participants cited a lack of appropriate patients with stroke and a preference for other treatment techniques as the main reasons for not having used MT in those 3 months. The participants’ responses are provided in Table 2.

Table 2.

Collected Comments from Participants at 3-Month Follow-Up

MT not appropriate for patients treated	Preference for other therapy	Other challenges	Positive comments
Did not want to use it too often, because patients already walking and working on functional-dynamic exercises. (P1)	Prefer to do what I’m comfortable with. Limited time, chose other treatments. (P5)	Difficult to incorporate into goal-based therapy. (P4)	Appreciated module as it helped visualize how to use MT for LE. (P6)
Current patients seemed too high functioning for MT, but would use with lower-level patient. (P6)	Practising in acute setting, had higher priorities, and time limited ability to use MT. (P7)	No appropriate mirror in treatment area. (P7)	Found it was a useful tool, would like to know how to facilitate buy-in with disinterested patient. (P2)
Acute patients had other higher priority issues, and in current state lacked attention and cognition to use MT. (P7)	—	Working on other units, less frequently working with patients with stroke. (P2, P8, P9)	Used it in long-sitting. It is a useful tool. Hard part is finding patient with long enough attention span. (P3)

Note: Dash indicates no applicable comment.

MT = mirror therapy; P = participant; LE = lower extremity.

Discussion

This is the first study to evaluate a KT intervention designed to encourage physical therapists’ implementation of MT for LE rehabilitation after stroke. The educational module significantly increased the participants’ perceived knowledge of, confidence in using, and willingness to implement MT. Three months after the intervention, 33% of the participants had used the technique in their practice. Although this is a minority, the finding that a 1-hour educational module changed practice among one-third of the participants is notable given the minimal resources behind this initiative, with no additional reminders or prompting.

The educational module was successful at increasing the participants’ perceived knowledge of MT, likely as a direct result of exposure to the information it provided. This is a promising finding because it highlights therapists’ ability to absorb a condensed version of the evidence regarding a topic in a short period of time. Moreover, it demonstrates the potential value of short in-service modules when increasing knowledge, rather than implementing a novel intervention, is the desired outcome. For example, short educational modules may be useful to update therapists’ knowledge base on already-used techniques, to reinforce EBP.²¹ It is important to note that we assessed perceived knowledge; incorporation of a quiz before and after the module would have confirmed an actual change in knowledge of MT.

The participants’ confidence in administering MT likely increased as a result of their enhanced knowledge of the subject and engagement in supervised hands-on practice after participating in a clinical demonstration of delivering MT. A similar implementation study by Levac and colleagues on rehabilitation using virtual reality also found a meaningful increase in confidence after delivering didactic information online over the course of multiple sessions.²⁷ Similarly, a KT study evaluating the impact of multiple educational modules in diabetes care found significant increases in health care professionals’ knowledge, perceived ability, and willingness to change their professional behaviour.²⁶

Building on these findings, our study suggests that, in the case of a much simpler therapy such as MT, a single in-person educational session may produce comparable increases in confidence and that developing in-depth modules requiring more time and resources may not be necessary. However, these findings of increased confidence in implementing MT must be considered in light of the fact that six of the participants did not use MT in the following 3 months. Although only one participant reported a lack of comfort with the technique as a reason for not using it, we cannot rule out that other participants similarly avoided MT because of a lack of confidence in their ability to properly deliver the treatment or identify an appropriate patient.

Our study also found that willingness to perform MT increased significantly, likely as a result of the increased knowledge of and confidence in performing MT that participants had gained from the educational session. Willingness to use MT was high before the intervention, which may suggest that the barrier to implementing it is not a lack of interest in adopting the novel treatment but rather a knowledge-to-action gap. This possibility highlights the notable lack of KT resources available to support clinicians in implementing new evidence-based therapies such as MT for LE hemiparesis.

The 3-month follow-up data show that even though the post-intervention questionnaire indicated that the participants’ willingness to use MT had increased after the intervention, their actual adoption of the technique was inconsistent: only three of the nine participants were using it. This points to the challenge of inspiring change in actual behaviour even after a technique is taught; Levac and colleagues also found that an increase in confidence with a technique did not lead to a change in therapists’ practice.²⁷ However, it is noteworthy that one-third of the practitioners had used MT at 3 months, especially given the small amount of time and resources needed to implement this KT intervention and that no further initiatives such as reminders or prompts were used. Indeed, along with interactive education sessions such as this one, regular reminders have consistently been identified as an effective strategy for promoting behavioural change among health professionals;²⁸ integrating reminders into the implementation strategy may benefit future KT interventions designed to increase the adoption of MT.

The 3-month follow-up interview found that identifying an appropriate patient for MT and developing recovery-stage-appropriate goals were the most common barriers to implementing MT among the participants, despite the suggestion in the research literature that MT can be effective in all stages of stroke recovery. This finding suggests that the 1-hour intervention may not have sufficiently addressed the challenges that a therapist may encounter in identifying appropriate patients and developing goals and treatment plans. To address this issue, future KT interventions aimed at increasing the use of MT for LE hemiparesis should expand on these topics. Preference for more familiar therapies was another notable reason why some participants did not implement MT. Again, implementing reminders and prompts may have encouraged the therapists to trial MT with their patients in the weeks after the module, which may also have developed their comfort in using the technique.

Future KT studies of behavioural change after an educational module should include tests of knowledge and practical skills to measure true change in knowledge and confidence, as well as audits to monitor change in practice. We found a ceiling effect for the willingness scores in our study; stipulating a threshold of baseline knowledge, confidence, or willingness should be considered when establishing inclusion criteria. In addition, incorporating follow-up prompts and providing specific equipment after the educational module is finished may improve actual behavioural change. For example, we gave a free-standing mirror to the two facilities at which the module was presented; this may have facilitated participants’ implementation of this novel treatment. Finally, including a comparison group, whether in the form of an active control or delayed intervention group, would ultimately confirm the effectiveness of such modules to change physiotherapy practice.

We acknowledge several limitations of this study. First, our sample size was small because of the limited number of physiotherapists working in stroke recovery in the Vancouver area. Second, the physiotherapists who volunteered to participate in the study were likely already aware of and favourable to the idea of using MT for the LE, given their high score on the pre-test questions related to willingness to use the therapy. Third, an element of acquiescence bias may have led participants to respond more positively to the post-test to satisfy a predicted research outcome. The study also assessed the participants’ perceived changes in knowledge and attitudes, which may not reflect an actual change. Finally, the single-group pre–post study design limits the interpretation of causality or effectiveness in this study.

Conclusion

This study provides preliminary evidence that a 1-hour education session can significantly improve therapists’ perceived knowledge of, confidence in, and willingness to use MT for LE stroke rehabilitation. It also suggests that such a session may lead to a change in practice at 3 months: some therapists (33%) had integrated the technique into their practice. Supplementing the module with follow-up reminders and prompts to use MT in the weeks after the intervention may increase implementation.

The ability of a 1-hour educational module to favorably influence physiotherapists’ knowledge about and attitudes toward an evidence-supported therapy adds to the body of knowledge in the field of KT. Simple, easy-to-administer treatment modalities that require minimal equipment may not necessitate multi-day courses or extensive online modules to increase understanding. Indeed, when investigating new therapeutic modalities, researchers should consider the knowledge-to-action gap. It may be worth directing increased research efforts toward easily understood therapies because they can be more rapidly taught to and implemented by clinicians.

Key Messages

What is already known on this topic

Knowledge translation (KT) and implementation studies have found that physical therapists cite a lack of time to consume and integrate research, as well as a lack of confidence, as significant barriers to implementing novel evidence-based therapies. Mirror therapy (MT) for lower-extremity (LE) paresis is an example of an evidence-based therapy that has yet to be widely considered or used.

What this study adds

This study demonstrates that an hour-long KT intervention can increase physical therapists’ understanding of, confidence in performing, and willingness to use MT to treat hemiparesis in the LE. In addition, a single intervention can create a change in practice; one-third of the participants in this study incorporated the technique into their practice.

APPENDIX 1: Educational Module Outline (using Boppps Lesson Format)

Component and time allotted	Topics
Completion of pre-questionnaire and socio-demographic survey (3 min: 0:00–3:00)	—
1. Introduction and learning objectives (3 min: 3:00-6:00)	(a) Bridge: stroke and LE hemiparesis
	(b) Pre-test: what is MT?
	(c) Learning objectives
2. Background of MT (participatory learning session) (3 min: 6:00–9:00)	(a) History: when and how it is used in practice
	(b) Neurophysiology of proposed mechanisms
	(c) Opportunity to practise illusion of MT with mirrors
3. Evidence for the use of MT with the LE (5 min: 9:00-14:00)	(a) Presentation of randomized controlled trials or systematic reviews; give examples of findings
	(b) Overall benefits and risks
	(c) Unilateral vs. bilateral
4. How to perform/teach MT (6 min: 14:00-20:00)	(a) When and when not to use it: patient selection (inclusion-exclusion criteria)
	(b) Set-up: equipment and environment
	(c) Demonstration of how to teach and cue a patient
	(d) Parameters: FITT
5. Case studies and hands-on skills practice (22 min: 20:00–42:00)	(a) Variety of mirrors available so therapists can experiment
	(b) Work through two case studies with role-play as patient and therapist (post-assessment)
6. Time for questions and summary (5 min: 42:00-47:00)	—
7. Completion of post-questionnaire (3 min: 47:00-50:00)	—

Note: Dashes indicate no applicable topic.

BOPPPS = Bridge, learning Objectives, Pre-test, Participatory learning session, Post-assessment, and Summary; LE = lower extremity; MT = mirror therapy; FITT = frequency, intensity, time, type.

APPENDIX 2: Questionnaire Items and Represented Domain

Item	Strongly disagree	Disagree	Neutral	Agree	Strongly agree
I am familiar with the current state of evidence for MT in neurorehabilitation. (knowledge)	❑	❑	❑	❑	❑
I can describe how MT works at the neurophysiological level. (knowledge)	❑	❑	❑	❑	❑
I can clinically reason the appropriate therapeutic prescription parameters for MT. (knowledge)	❑	❑	❑	❑	❑
I would be able to successfully and safely set up the environment to administer MT. (confidence)	❑	❑	❑	❑	❑
I would be able to teach MT techniques for the lower extremity to a patient with proper cueing. (confidence)	❑	❑	❑	❑	❑
I would be able to educate a patient on how MT works and how it could benefit them. (confidence)	❑	❑	❑	❑	❑
I believe that patients could benefit from MT for the treatment of lower-extremity hemiparesis. (willingness)	❑	❑	❑	❑	❑
I believe that MT is an easy-to-administer therapy for lower-extremity hemiparesis. (willingness)	❑	❑	❑	❑	❑
I would consider using MT in my future clinical practice with a stroke patient. (willingness)	❑	❑	❑	❑	❑

MT = mirror therapy.