Stroke is a common global healthcare problem that is serious and disabling (Warlow 2008). Around 15 million people worldwide suffer from a stroke every year (WHO 2003), and the worldwide prevalence in the general population is estimated to be 0.7%. Stroke afflicts people of all ages. Post-stroke deficits can persist long-term despite a rapid rate of motor recovery that commonly occurs in the first five to six weeks. Normal walking patterns become impaired in people with stroke, primarily caused by muscle weakness, spasticity, impaired sensorimotor control or loss of cognitive function. Secondary impairments, such as muscle atrophy and reduced aerobic capacity, lead to further gait impairment (Macko 2001; Van Nunen 2012). The biggest effect on people who have had a stroke, and their families, is usually caused by long-term impairment, limitations of activities and reduced participation (Langhorne 2009).
Description of the condition
Most stroke survivors regain the ability to walk; however, the gait pattern may be deficient (Yelnik 1999). Walking after stroke has been identified as the top priority for rehabilitation (Chan 1997). People who have had a stroke have a low level of ambulatory activity (Michael 2005). Hemiparesis is one of the most common post-stroke impairments that contributes to reduced gait performance. It has been noted in the literature that people with hemiparesis walk significantly slower than healthy persons and after six months reach only 40% to 50% of the walking distance of age-matched healthy people (Pohl 2002). The pattern of walking deviates due to many internal factors, such as inadequate or abnormal muscle recruitment or muscle weakness leading to an inability to initiate or control joint movements. Paralysis hinders foot positioning and loading when standing (Lee 2005). The hemiparetic foot frequently adopts a plantar-flexed position, not only in the swing phase but often throughout the stance phase as well. This may be due to the presence of increased plantar flexor tone, inappropriate plantar flexor activity, plantar flexor contracture or dorsiflexor weakness. Regardless of its cause the plantar flexion results in lack of weight bearing on the heel. As a further consequence the excessive plantar flexion resists forward rolling of the tibia over the ankle joint leading to knee hyperextension, therefore the ground reaction force (GRF) passes in front of the knee leading to instability. In addition, the plantar-flexed position pulls the GRF in front of the hip causing an excessive flexion moment of the hip during the late stages of the stance phase (Meadows 2008). It is clearly stated in the literature that hemiparesis induces ankle control disturbances and equinovarus deformity, leading to difficulty in walking and an increased risk of falling (Abe 2009).
Description of the intervention
One of the most common ways to manage motor impairments and to minimise gait deviation is the use of externally applied devices. Such devices, known as orthoses, are used to modify the structural and functional characteristics of the neuromusculoskeletal system (Ponton 1997). To promote walking ability, ankle foot orthoses (AFOs) are frequently prescribed to various groups of people who experience loss of control or impairments of muscle function around the ankle. AFOs are clinical devices designed to improve walking ability in the absence of natural substitutive patterns (Leung 2002; Michael 2008). It has been suggested that they improve the dynamic efficiency of gait, that is the degree to which the gait is well controlled and energy efficient. They can be designed with sufficient mechanical lever arms to control the ankle complex directly and to influence the knee joint indirectly (Michael 2008). There are many types of AFOs that may vary in their biomechanical designs and which are prescribed to people with hemiparesis. Broadly, they can be classified into prefabricated and custom-made orthoses. Prefabricated AFOs are usually made of plastic. There are a number of prefabricated designs available but the most common design is the posterior leaf spring. The benefits of prefabricated orthoses are few and they are only used for improving the swing phase of walking. They are often used temporarily, such as during early mobilisation before a custom-made orthosis can be made available. Custom-made orthoses are usually prescribed for more complex gait abnormalities associated with stroke. These AFOs are most appropriate for controlling significant ankle triplanar deformity and when knee or hip problems are present (Condie 2004; Condie 2008). There are many different designs of custom-made AFOs, such as:
- the posterior leaf spring AFO;
- the solid AFO;
- the hinged or articulated AFO;
- the dorsiflexion assist AFO;
- the plantar flexion stop AFO; and
- the energy return or ground reaction AFO.
The prescription of the various types of custom-made orthoses should be based on the deficits experienced by each individual and the goals they need to achieve. When properly designed and fitted, it is claimed that custom-made orthoses have more beneficial effects during the stance and swing phases of gait. Prefabricated and custom-made orthoses are very commonly used in the clinical set-up. However, custom-made orthoses are more commonly used in developed countries, because of the associated costs. The function of prefabricated orthoses and custom-made orthoses differs significantly. An optimally designed and fitted custom-made orthosis would be extremely beneficial (Condie 2004).
How the intervention might work
Provided they are adequately stiff, AFOs can prevent plantar flexion of the foot in the swing phase of gait and improve ground clearance, reducing the risk of falling. This is achieved by the three-point force system applied to the posterior calf, plantar surface of the foot near the metatarsal heads, and the dorsum of the foot near the ankle joint, which maintains the ankle in the proper position (Meadows 2008). AFOs not only alter the biomechanics at the ankle joint, they also indirectly change the hip and knee alignment by realigning the tibia to a more normal position of approximately 10° forward inclination; the weight bearing on the foot is redistributed to the entire plantar surface rather than the lateral aspect of the foot and this shifts the GRF to the posterior. During the stance phase, the GRF is positioned posterior to the knee thereby positively influencing the hip and knee stability. As the GRF is closer to the knee joint, it prevents hyperextension during the late stance phase. The alteration of the GRF positioning further changes the alignment of the hip joint to the anterior; this helps in reducing the abnormal flexion movement at the hip during the terminal stance. This adaptation of the GRF on the hip and knee joints normalises the demand on the hemiparetic limb by improving both control and stability during walking (Meadows 2008). However, it has been observed that these effects are based on theoretical assumptions and are not consistently demonstrated among people using AFOs. Clinical studies indicate that symmetry of gait improves with the use of AFOs compared with walking without orthoses. Various studies suggest that AFOs enhance weight bearing through the affected limb during both walking and standing (Pohl 2006; Wang 2005). Some studies indicate that AFOs can control knee recurvatum during stance (Boudarham 2013; Tyson 2013). It can be said that these effects decrease the biomechanical challenges during gait thereby reducing the required neuromuscular response and improving the mobility of people suffering from stroke.
Why it is important to do this review
Despite the fact that AFOs are thought to have beneficial effects on functional walking ability, there is no conclusive evidence to endorse or refute orthotic treatment around the ankle in the context of promoting walking ability (Harlaar 2010). Several studies present conflicting results, or even mixed results within one study (Brehm 2008; Buckon 2004; Churchill 2003; De 2004; Desloovere 2006; Leung 2002; Radtka 2005). The International Society of Prosthetics and Orthotics has organised consensus meetings on the treatment of various disabling conditions such as cerebral palsy, stroke, and poliomyelitis. However, the resulting consensus reports are still inconclusive regarding specific prescription guidelines for orthotic treatment of ankles (Condie 1995; Condie 2003; Heim 1997; Hijmans 2006; Morris 2008). There has been one narrative review exploring the impact of AFOs on gait and leg muscle activity in people with hemiplegia. In this review it was found that the evidence for the impact of an AFO on muscle activity in the paretic leg was weak, although there may be immediate kinematic and temporal improvements (Leung 2002). One recently published systematic review has assessed the effectiveness of AFOs on mobility and balance in people with stroke; the available evidence suggests that an AFO can improve walking and balance immediately, but the effects and acceptability of long-term usage need to be evaluated (Tyson 2013a). Many clinicians believe that wearing an AFO has a negative effect on the mechanics of walking that may outweigh the benefits. The concerns include probable deviation from the normal kinematics of gait and muscle disuse, particularly of the ankle dorsiflexors (Geboers 2001; Geboers 2002). It is important to ascertain the effectiveness of AFOs because regaining safe independent mobility is the primary goal for stroke patients.
To assess the effects of AFOs on walking ability (speed, velocity, temporal-spatial gait parameters) in people with stroke.
Criteria for considering studies for this review
Types of studies
We will only include published and unpublished randomised controlled trials (RCTs) of either parallel or crossover design. For crossover studies, we will only use the data from the first phase.
Types of participants
All adults and children (over two years of age) with a confirmed diagnosis of stroke. A diagnosis of stroke fulfils the clinical criteria of the World Health Organization (WHO) definition, of "rapidly developed clinical signs of focal/global disturbances of cerebral flow clinically lasting for more than 24 hours or leading to death with no other apparent case of vascular origin" (Hatano 1976). A diagnosis of stroke encompasses ischaemic and haemorrhagic stroke (including subarachnoid, intraventricular or intracerebral haemorrhage). We will not include trials of children with cerebral palsy. We will exclude patients with progressive brain and spinal pathologies.
Types of interventions
This review will include studies where an AFO is used to improve walking, functional mobility or activity limitations, or which report any incidence of adverse events in participants with stroke. We will include studies that involved other co-interventions (for example electrical stimulation) provided the co-interventions were applied in the same manner to both the control and experimental group participants. We will exclude interventions that are not specifically AFOs, for example taping, strapping, compressive bandages, air splinting, casting or any shoe or footwear modifications. We will also exclude AFOs that are part of functional electrical stimulation.
We will include the following comparisons:
- AFO versus no AFO;
- AFO with another intervention versus no AFO with another intervention.
Types of outcome measures
We will include trials in which the outcomes could be assessed either with or without the AFOs.
Spatial and temporal parameters of walking. This includes walking speed, velocity, cadence, stride time, stride length, and step length.
- Functional mobility: Timed Up and Go Test (Podsiadlo 1991), Dynamic Gait Index.
- Fall frequency.
- Participation restrictions and impact on carers, quality of life (QoL): this can include measures reflecting return to work, leisure or home life.
- Energy expenditure: this can include oxygen consumption or metabolic equivalent (Ivey 2006).
- Follow-up: long-term effects of outcome on patient satisfaction following the intervention.
- Compliance with the orthosis or duration of use of the orthosis.
- Adverse events, such as skin damage, pain, discomfort (present or absent).
Search methods for identification of studies
See the 'Specialized Register' section in the Cochrane Stroke Group module. We will search for trials in all languages and arrange for the translation of relevant articles where necessary.
We will search the trials registers of the Cochrane Stroke Group and the Cochrane Musculoskeletal Group. We will also search the following electronic databases to identify potential studies:
- Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library, latest issue);
- MEDLINE (from 1948) (Appendix 1);
- EMBASE (from 1980);
- CINAHL (from 1982);
- AMED (from 1985);
- Physiotherapy Evidence Database (PEDro) (www.pedro.org.au/);
- REHABDATA (www.naric.com/research/rehab/);
- RECAL database (comprehensive database in the field of prosthetics, orthotics and related physical medicine and rehabilitation) (http://cdlr.strath.ac.uk/recal/).
We will adapt the MEDLINE search strategy (Appendix 1), as appropriate, for each database.
We will search for ongoing trials in the following trials registers:
- Stroke Trials Registry (www.strokecenter.org/trials/);
- Current Controlled Trials (www.controlled-trials.com);
- WHO International Clinical Trials Registry Platform(http://apps.who.int/trialsearch/).
Searching other resources
- search reference lists of all retrieved articles and other reviews on the topic;
- use Science Citation Index Cited Reference Search for forward tracking of important articles;
- contact study authors, researchers and experts in the field;
- search Google Scholar (http://scholar.google.co);
- contact equipment manufacturers.
Data collection and analysis
We will collate the search results using bibliographic software and will remove duplicates before the screening process.
Selection of studies
Three authors (RP, SS, RK) will independently screen the results of the searches for relevant articles based on the title and abstract and classify them as either include, exclude or unclear. We will obtain the full-text articles for the included studies and also the unclear studies to determine their eligibility. If there are any disagreements between review authors, we will resolve them by discussion. We will seek translations for studies published in languages other than English.
Data extraction and management
All review authors, working independently, will extract the data using a specifically formulated pre-tested data extraction form. The extracted data will include characteristics of participants (age; gender; time post-stroke; hemisphere of stroke; type of stroke; modified Ashworth score; type of AFO used; duration of use of AFO; previous use of AFO; walking aids; baseline measures capturing mobility, temporal and spatial parameters of gait), information on study design (type of randomisation, type of concealment, number of participants), aspects of the intervention (details of the intervention and the control intervention, duration of intervention), outcome measures and dropouts. If any of the trials included in this review were authored by a member of our review team, the author will not extract the data from that study.
Assessment of risk of bias in included studies
All review authors will independently assess the methodological quality of the included RCTs. We will assess the risk of bias in included studies by using the risk of bias tool described in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). We will assess and rate the following methodological domains.
- Sequence generation
- Low risk of bias: using a computerised random generator, random number tables, coin tossing or any other valid method.
- High risk of bias: sequence generation and allocation done by invalid methods such as using odd or even date of birth, or allocation by the judgement of the clinician.
- Unclear risk of bias: insufficient information provided about the sequence generation process.
- Allocation sequence concealment
- Low risk of bias: allocation concealed so that neither the investigators or participants know group assignment at the time of study entry. Valid methods include central randomisation or serially numbered, opaque, sealed envelopes.
- High risk of bias: the method of allocation is not concealed (e.g. list of random numbers, unsealed or non-opaque envelopes) leading to transparency in group assignments thereby introducing selection bias.
- Unclear risk of bias: insufficient information provided about the concealed allocation process.
- Blinding of participants, personnel and outcome assessors
- Low risk of bias: either participants or some study key personnel could not or were not blinded but the outcome assessment was blinded and the non-blinding of others is unlikely to introduce bias.
- High risk of bias: no blinding or incomplete blinding, and the outcome measurement is likely to be influenced by lack of blinding.
- Unclear risk of bias: insufficient information or the study did not mention it.
- Incomplete outcome data
- Low risk of bias: missing data have been imputed using appropriate methods such as intention-to-treat analysis.
- High risk of bias: authors did not impute intention-to-treat analysis for missing data.
- Unclear risk of bias: insufficient reporting of attrition and exclusions, no reasons for missing data provided.
- Selective outcome reporting
- Low risk of bias: published article reports primary and secondary outcomes that are of interest to the review in the pre-specified way.
- High risk of bias: study did not report the pre-specified primary outcome.
- Unclear risk of bias: insufficient information to permit judgement of yes or no.
- Other potential threats to validity
- Low risk of bias: the study appears to be free of other sources of bias.
- High risk of bias: there is bias pertaining to the study design (e.g. extreme baseline imbalance).
- Unclear risk of bias: insufficient information to assess whether any important risk of bias exists.
Review authors will not assess the risk of bias in studies in which they were involved. Studies published in languages other than English will be assessed for risk of bias by authors fluent in that language.
Measures of treatment effect
We will derive pooled mean differences (MDs) and 95% confidence intervals (CIs) for continuous outcomes using the same scale of measurement. We will calculate pooled standardised mean difference (SMD) for continuous outcomes using different scales of measurement (for example angular or linear measurements of joint range of motion). For clinical interpretation we will convert these pooled SMDs to the most commonly used original scales in that comparison by using the formula stated in Chapter 12 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). We will express dichotomous outcomes as relative risk (RR) and 95% CI.
Unit of analysis issues
For crossover studies the major concern is the carry-over effect. For this review we intend to use the first phase data from the included studies. If data for the first phase are not reported separately we will contact the study authors requesting them to provide the data on the first phase. To incorporate these differences that may lead to spurious heterogeneity we will perform a subgroup analysis based on the study design (parallel group trials and crossover trials).
Dealing with missing data
We will contact study authors regarding any missing data. If the authors are unavailable or the additional data are still insufficient for analysis, we will provide a description of the study in the review. We will use an intention-to-treat analysis. When there are more than 10% dropouts we will assess the robustness of the assumption by performing sensitivity analyses.
Assessment of heterogeneity
We will perform a visual inspection of the forest plots to see if the CIs for all the studies are overlapping. We will also use the Chi² statistic, along with the I² statistic, to quantify the amount of heterogeneity. We will classify an I² value of more than 50% as substantial and we will explore the possible causes of the heterogeneity by performing subgroup analyses. We will use a random-effects model if the I² is greater than 50%. If there is a low level of statistical heterogeneity (that is I² less than 50%), we will combine studies using a fixed-effect model.
Assessment of reporting biases
If we include more than 10 studies in a meta analysis we will perform funnel plot analysis to assess reporting bias.
We will perform data analysis using RevMan 5.3 (RevMan 2014) following the guidance provided in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). One review author will verify all the extracted data before entering them for analysis. We will calculate the RR with 95% CI and SMD or MD with 95% CI for dichotomous and continuous outcomes.
Subgroup analysis and investigation of heterogeneity
We will examine the following with the use of subgroup analysis.
- Effect of AFO when used for varying periods of time (immediate effects and long-term effects).
- Effect of AFO when used by participants of different ages.
- Effect of various types of AFO when used by participants.
- All outcome measures with and without AFO.
We will perform sensitivity analyses to assess the robustness of the findings by excluding the studies from the analysis which are at high risk of bias (measured by both allocation concealment and blinding) and also that have higher rates of dropouts (> 15% of participants).
Grading the quality of the evidence
We will use the GRADE approach to evaluate the quality of the evidence of the trials included in this review (GRADE Working Group 2004; Guyatt 2008; Guyatt 2008a; Schunemann 2006). The GRADE approach specifies four levels of quality:
- high quality, randomised trials;
- medium quality, downgraded randomised trials;
- low quality, double-downgraded randomised trials; and
- very low quality, triple-downgraded randomised trials.
The quality of evidence will be downgraded if:
- there are limitations in the design and implementation of available studies suggesting high likelihood of bias;
- there is indirectness of evidence (indirect population, intervention, control, outcomes);
- there is unexplained heterogeneity or inconsistency of results (including problems with subgroup analyses);
- there is imprecision of results (wide CIs); and
- there is a high probability of publication bias.
The quality of evidence will be upgraded if:
- there is a large magnitude of effect;
- all plausible confounding reduces a demonstrated effect or suggests a spurious effect in studies which show no effect; and
- there is a dose-response gradient.
Summary of findings tables
We will compile a summary of findings table using GRADEpro (GRADEpro 2008) to present the main findings of the review in a transparent and simple format. The summary of findings table will include the following outcomes: walking, functional mobility, fall frequency, activity limitations, energy expenditure, follow-up outcome, and adverse events.
We acknowledge the South Asian Cochrane Centre for supporting this review and also the members of the Cochrane Stroke Group Editorial Board.
Appendix 1. MEDLINE search strategy
1. cerebrovascular disorders/ or exp basal ganglia cerebrovascular disease/ or exp brain ischemia/ or exp carotid artery diseases/ or exp intracranial arterial diseases/ or exp "intracranial embolism and thrombosis"/ or exp intracranial hemorrhages/ or stroke/ or exp brain infarction/ or stroke, lacunar/ or vasospasm, intracranial/ or vertebral artery dissection/
2. (stroke or poststroke or post-stroke or cerebrovasc$ or brain vasc$ or cerebral vasc$ or cva$ or apoplex$ or SAH).tw.
3. ((brain$ or cerebr$ or cerebell$ or intracran$ or intracerebral) adj5 (isch?emi$ or infarct$ or thrombo$ or emboli$ or occlus$)).tw.
4. ((brain$ or cerebr$ or cerebell$ or intracerebral or intracranial or subarachnoid) adj5 (haemorrhage$ or hemorrhage$ or haematoma$ or hematoma$ or bleed$)).tw.
5. hemiplegia/ or exp paresis/
6. (hemipleg$ or hemipar$ or paresis or paretic).tw.
7. exp gait disorders, neurologic/
8. 1 or 2 or 3 or 4 or 5 or 6 or 7
9. Foot Orthoses/
10. ((foot or ankle-foot or foot-drop) adj5 (orthos$ or orthotic or brace$ or splint$)).tw.
11. (AFO or AFOs).tw.
12. ((ankle or foot) adj3 support$).tw.
13. 9 or 10 or 11 or 12
14. orthotic devices/ or braces/
15. (orthosis or orthoses or orthotic or brace$ or splint$).tw.
16. 14 or 15
17. exp Lower Extremity/
18. foot joints/ or ankle joint/
19. (lower extremit$ or leg or legs or ankle$ or foot or feet or heel$ or toe$ or hip or knee or knees or thigh$).tw.
20. (walk$ or gait$ or ambulat$ or mobil$ or locomot$ or balanc$ or stride or foot-drop).tw.
21. gait/ or locomotion/ or exp walking/
22. 17 or 18 or 19 or 20 or 21
23. 16 and 22
24. 13 or 23
25. Randomized Controlled Trials as Topic/
26. random allocation/
27. Controlled Clinical Trials as Topic/
28. control groups/
29. clinical trials as topic/
30. double-blind method/
31. single-blind method/
33. placebo effect/
34. cross-over studies/
35. Therapies, Investigational/
36. Research Design/
37. randomized controlled trial.pt.
38. controlled clinical trial.pt.
39. clinical trial.pt.
40. (random$ or RCT or RCTs).tw.
41. (controlled adj5 (trial$ or stud$)).tw.
42. (clinical$ adj5 trial$).tw.
43. ((control or treatment or experiment$ or intervention) adj5 (group$ or subject$ or patient$)).tw.
44. (quasi-random$ or quasi random$ or pseudo-random$ or pseudo random$).tw.
45. ((control or experiment$ or conservative) adj5 (treatment or therapy or procedure or manage$)).tw.
46. ((singl$ or doubl$ or tripl$ or trebl$) adj5 (blind$ or mask$)).tw.
47. (cross-over or cross over or crossover).tw
48. (placebo$ or sham).tw.
50. (assign$ or allocat$).tw.
52. 8 and 24 and 51
53. exp animals/ not humans.sh.
54. 52 not 53
Contributions of authors
Conceiving the review: SS
Designing the review: SS, RP, RK
Co-ordinating the review: SS
Writing the protocol: SS, RP, RK
Providing general advice on the protocol: RP, BS, SS, JSP, RK
Declarations of interest
None of the authors have any conflicts of interest.
Sources of support
- No sources of support supplied
- Department for International Development (DFID), UK.Funding for the Prof. BV Moses Centre for Advanced Research in Evidence-Informed Healthcare; Salary support for Richard Kirubakaran through the Effective Healthcare Research Consortium.