Khan Et Al-2015-Cochrane Database of Systematic Reviews PDF

Cochrane
Library
Cochrane Database of Systematic Reviews

Telerehabilitation for persons with multiple sclerosis (Review)
Khan F, Amatya B, Kesselring J, Galea M
Khan F, Amatya B, Kesselring J, Galea M.

Telerehabilitation for persons with multiple sclerosis.
Cochrane Database of Systematic Reviews 2015, Issue 4. Art. No.: CD010508.
DOI: 10.1002/14651858.CD010508.pub2.
www.cochranelibrary.com

Telerehabilitation for persons with multiple sclerosis (Review)
Copyright © 2015 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.
Cochrane Trusted evidence.
Informed decisions.

Library Better health. Cochrane Database of Systematic Reviews
TABLE OF CONTENTS
HEADER......................................................................................................................................................................................................... 1
ABSTRACT..................................................................................................................................................................................................... 1
PLAIN LANGUAGE SUMMARY....................................................................................................................................................................... 2
SUMMARY OF FINDINGS.............................................................................................................................................................................. 3
BACKGROUND.............................................................................................................................................................................................. 8
OBJECTIVES.................................................................................................................................................................................................. 9
METHODS..................................................................................................................................................................................................... 9
RESULTS........................................................................................................................................................................................................ 12
Figure 1.................................................................................................................................................................................................. 13
Figure 2.................................................................................................................................................................................................. 15
Figure 3.................................................................................................................................................................................................. 16
DISCUSSION.................................................................................................................................................................................................. 19
AUTHORS' CONCLUSIONS........................................................................................................................................................................... 21
ACKNOWLEDGEMENTS................................................................................................................................................................................ 22
REFERENCES................................................................................................................................................................................................ 23
CHARACTERISTICS OF STUDIES.................................................................................................................................................................. 26
ADDITIONAL TABLES.................................................................................................................................................................................... 39
APPENDICES................................................................................................................................................................................................. 43
CONTRIBUTIONS OF AUTHORS................................................................................................................................................................... 43
DECLARATIONS OF INTEREST..................................................................................................................................................................... 43
SOURCES OF SUPPORT............................................................................................................................................................................... 44
DIFFERENCES BETWEEN PROTOCOL AND REVIEW.................................................................................................................................... 44
INDEX TERMS............................................................................................................................................................................................... 44
Telerehabilitation for persons with multiple sclerosis (Review) i

Informed decisions.

[Intervention Review]
Telerehabilitation for persons with multiple sclerosis
Fary Khan1,2,3, Bhasker Amatya1, Jurg Kesselring4, Mary Galea5
1Department of Rehabilitation Medicine, Royal Melbourne Hospital, Royal Park Campus, Melbourne, Australia. 2School of Public Health
and Preventive Medicine, Monash University, Melbourne, Australia. 3Department of Medicine, Dentistry & Health Sciences, University
of Melbourne, Melbourne, Australia. 4Department of Neurology and Neurorehabilitation, Rehabilitation Center, Valens Hospital, Valens,
Switzerland. 5Department of Medicine (Royal Melbourne Hospital), The University of Melbourne, Parkville, Australia
Contact address: Fary Khan, Department of Rehabilitation Medicine, Royal Melbourne Hospital, Royal Park Campus, Poplar Road,
Parkville, Melbourne, Victoria, 3052, Australia. [email protected].
Editorial group: Cochrane Multiple Sclerosis and Rare Diseases of the CNS Group
Publication status and date: New, published in Issue 4, 2015.
Citation: Khan F, Amatya B, Kesselring J, Galea M. Telerehabilitation for persons with multiple sclerosis. Cochrane Database of
Systematic Reviews 2015, Issue 4. Art. No.: CD010508. DOI: 10.1002/14651858.CD010508.pub2.
ABSTRACT
Background
Telerehabilitation, an emerging method, extends rehabilitative care beyond the hospital, and facilitates multifaceted, often
psychotherapeutic approaches to modern management of patients using telecommunication technology at home or in the community.
Although a wide range of telerehabilitation interventions are trialed in persons with multiple sclerosis (pwMS), evidence for their
effectiveness is unclear.
Objectives
To investigate the effectiveness and safety of telerehabilitation intervention in pwMS for improved patient outcomes. Specifically,
this review addresses the following questions: does telerehabilitation achieve better outcomes compared with traditional face-to-face
intervention; and what types of telerehabilitation interventions are effective, in which setting and influence which specific outcomes
(impairment, activity limitation and participation)?
Search methods
We performed a literature search using the Cochrane Multiple Sclerosis and Rare Diseases of the Central Nervous System Review Group
Specialised Register( 9 July, 2014.) We handsearched the relevant journals and screened the reference lists of identified studies, and
contacted authors for additional data.
Selection criteria
Randomised controlled trials (RCTs) and controlled clinical trials (CCTs) that reported telerehabilitation intervention/s in pwMS and
compared them with some form of control intervention (such as lower level or different types of intervention, minimal intervention,
waiting-list controls or no treatment (or usual care); interventions given in different settings) in adults with MS.
Data collection and analysis

Two review authors independently selected studies and extracted data. Three review authors assessed the methodological quality of
studies using the GRADEpro software (GRADEpro 2008) for best-evidence synthesis. A meta-analysis was not possible due to marked
methodological, clinical and statistical heterogeneity between included trials and between measurement tools used. Hence, we performed
a best-evidence synthesis using a qualitative analysis.
Telerehabilitation for persons with multiple sclerosis (Review) 1

Informed decisions.

Main results
Nine RCTs, one with two reports, (N = 531 participants, 469 included in analyses) investigated a variety of telerehabilitation interventions
in adults with MS. The mean age of participants varied from 41 to 52 years (mean 46.5 years) and mean years since diagnosis from 7.7 to
19.0 years (mean 12.3 years). The majority of the participants were women (proportion ranging from 56% to 87%, mean 74%) and with a
relapsing-remitting course of MS. These interventions were complex, with more than one rehabilitation component and included physical
activity, educational, behavioural and symptom management programmes.
All studies scored 'low' on the methodological quality assessment. Overall, the review found 'low-level' evidence for telerehabilitation
interventions in reducing short-term disability and symptoms such as fatigue. There was also 'low-level' evidence supporting
telerehabilitation in the longer term for improved functional activities, impairments (such as fatigue, pain, insomnia); and participation
measured by quality of life and psychological outcomes. There were limited data on process evaluation (participants'/therapists'
satisfaction) and no data available for cost effectiveness. There were no adverse events reported as a result of telerehabilitation
interventions.
Authors' conclusions
There is currently limited evidence on the efficacy of telerehabilitation in improving functional activities, fatigue and quality of life in adults
with MS. A range of telerehabilitation interventions might be an alternative method of delivering services in MS populations. There is
insufficient evidence to support on what types of telerehabilitation interventions are effective, and in which setting. More robust trials are
needed to build evidence for the clinical and cost effectiveness of these interventions.
PLAIN LANGUAGE SUMMARY
Review questions
Does telerehabilitation achieve better outcomes in persons with multiple sclerosis compared with traditional face-to-face intervention?
What types of telerehabilitation interventions are effective, in which setting and influence which specific outcomes?
Background
Multiple sclerosis (MS) is a common disease of the nervous system among young adults, with no cure and causing long-term disability.
Rehabilitation provides treatments and therapies to lessen the impact of any disability and improve function. Despite recent advances in
MS care including rehabilitation, many people with MS are unable to access these developments due to limited mobility, fatigue and related
issues, and costs associated with travel.Telerehabilitation is a newer approach to delivering rehabilitation programmes at the patient’s
home or in the community, using telecommunication technology such as phone lines, video technology, internet applications and others.
A wide range of telerehabilitation interventions are trialed in persons with multiple sclerosis, however, evidence for their effectiveness is
still unclear.
Study characteristics
This review looked for evidence on how telerehabilitation interventions work in adults with MS. We searched widely for randomised
controlled trials (RCTs), a particular kind of study where participants are placed in treatment groups by chance (that is, randomly) because
in most settings these provide the highest quality evidence. We were interested in studies that compared a telerehabilitation programme
with standard or minimal care, or with different kinds of rehabilitation programmes.
Key results
We found nine relevant RCTs covering 531 participants (469 included in the analyses), evaluating a wide variety of telerehabilitation
interventions in persons with MS. The telerehabilitation interventions evaluated were complex, with more than one rehabilitation
component and included physical activity, educational, behavioural and symptom management programmes. These interventions had
different purposes and used different technologies, so a single overall definite conclusion was not possible. The methodological quality
of the included studies is low and varied among the studies.
Quality of evidence
There was 'low-quality' evidence from the included RCTs to support the benefit of telerehabilitation in reducing short-term disability and
managing symptoms such as fatigue in adults with MS. We found limited evidence to support the benefit of telerehabilitation interventions
in improving disability, reducing symptoms and improving quality of life in the longer term. Furthermore, the interventions and outcomes
being investigated in the included studies were different to each other. No studies reported any serious harm from telerehabilitation and
there was no information on the associated costs.
There is a need for further research to assess the effects of the range of telerehabilitation techniques and to establish the clinical and cost
effectiveness of these interventions in people with MS. The evidence in this review is up to date to July 2014.
Informed decisions.

SUMMARY OF FINDINGS

Summary of findings for the main comparison.
Patient or population: People with multiple sclerosis
Settings: Participants' home, MS regional centres
Intervention: Telerehabilitation
Comparison: Standard care in rehabilitation centres, participants in wait-list, other type/intensity of rehabilitation
intervention
Outcomes No of Partic- Effect of telerehabilitation interventions for people with multiple sclerosis Quality of
ipants the evi-
(studies) dence
(GRADE) #
Change in functional activity
Change in 232 (inter- Two studies (Dlugonski 2012; Motl 2011, N = 99) with same cohort of participants ⊕⊕⊝⊝
disability di- vention showed significant improvement in physical activity in the treatment group at post- low1
rectly post- group = 122) intervention assessment as measured by GLTEQ (P < 0.01). Weekly step count (pe-
interven- (6 studies) dometer) increased significantly in the treatment group at post-intervention assess-
tion ment (P < 0.001)
Measures:
GLTEQ, One study (Frevel 2014, N = 18) showed significant improvement in dynamic and
DGI, BBS, static balance capacity compared to baseline values in both intervention group (e-
ARAT, NHPT, training) (DGI: P = 0.016, BBS: P = 0.011) and control (hippotherapy) group (DGI: P =
25FWT, CES, 0.011, BBS: P = 0.011). There was no difference between groups
VPR
One study (Huijgen 2008, N = 35) showed no statistically significant differences be-
Follow-up:
tween intervention and control groups in arm function as measured by ARAT (mean
depend-
change 1.26, 90% CI -1.90 to 4.42) and NHPT (mean change 7.24, 90% CI -6.55 to
ed on the
23.25)
type of in-
tervention; One study (Paul 2014, N = 30) showed that gait speed measured using 25FWT in-
range from creased in the intervention group compared to the control group but this was not
(1 month – statistically significant (P = 0.170); and the intervention group showed a statistically
12 weeks) significant improvement in the physical subscale of the MSIS (P = 0.048)
One study (Gutíerrez 2013a, N = 50) showed improvements in balance and pos-
tural control, with a significant increase in CES of the intervention group (mean
change; 8.21 points, P < 0.001), but no significant improvement in the control group
(mean change: 1.93, P = 0.123). Visual Preference Ratio (VPR) and the contribution
of vestibular information (Vestibular Ratio) improved significantly in the interven-
tion group (P < 0.001), but not in the control group (P > 0.05). There were significant
post-treatment differences between treatment and control groups in the CES (F =
37.873, P < 0.001) and the VPR (F = 12.156, P < 0.001). Significant post-treatment dif-
ferences between groups were also found for the ability to accept incorrect visual
information expressed by the visual conflict parameter (F = 15.05, P < 0.000). There
were no significant between-group differences in the contribution of the visual sys-
tem (F = 2.64, P = 0.11) or use of somatosensory information (F = 0.117, P = 0.734) in
the maintenance of balance and stability
Change in 45 (interven- One study (Dlugonski 2012, N = 45) reported that the treatment group showed a sig- ⊕⊕⊝⊝
short-term tion group = nificant increase in physical activity at 3-month follow-up compared to the control low2
disability 3 22) group as measured by GLTEQ (P < 0.001). There was a non-significant change in as-
months or (1 study) sessment scores from post-intervention to 3-month follow-up (P = 0.61)

Informed decisions.

less after
the start of
the inter-
vention
Measures:
GLTEQ
Follow-up:
up to 3
months
Change in 82 (interven- One study with 2 reports (Pilutti 2014, N = 82) showed a significant and positive ef- ⊕⊕⊝⊝
long-term tion group = fect of the intervention on increase in 6MWT distance relative to those in the control low2
disability 41) group (P = 0.07). Physical activity increased most in those with mild disability in the
more than 3 (1 study with intervention group.
months af- 2 reports)
ter the in-
tervention
Mea-
sure:6MWT
Follow-up: 6
months – 2
years
Change in symptoms or impairments
Change 265 (inter- One study (Finlayson 2011, N = 190) showed a significant reduction in fatigue in in- ⊕⊕⊝⊝
in impair- vention tervention group compared to a wait-list control group immediately after interven- low3
ments di- group = 138) tion as measured by FIS sub-scales (Mean (SD): Cognitive -3.12 (6.1), P = 0.001; Phys-
rectly post- (4 studies) ical -2.53 (6.4), P = 0.014; Social -6.01 (12.1), P = 0.002)
interven-
tion One study (Egner 2003, N = 27) reported similar fatigue scores (measured using FSS)
Measures: for all 3 groups (video, telephone and standard care) at 9 weeks post-intervention;
FIS, FSS, however the video group had significantly lower scores than the other 2 groups at
MFIS, MS month 6 (P < 0.05; telephone: SE = 0.478; standard care: SE = 0.536) and month 18 (P
Symptom < 0.05; telephone: SE = 0.569; standard care: SE = 0.624)
Cheklist
One study (Frevel 2014, N = 18) reported that fatigue improved significantly in the
Follow-up: control (hippotherapy) group (P < 0.05 for all MFIS subscales); while the e-training
depend- group improved only on the MFIS cognitive subscale (P = 0.031). A significant differ-
ed on the ence between the groups was noted only in the cognitive subscale of the MFIS ( P =
type of in- 0.012)
tervention;
One study (Paul 2014, N = 30) reported no improvements in symptoms as measured
range from
by MS Symptom Checklist.
(1 month –
12 weeks)
Change 190 (inter- One study (Finlayson 2011, N = 190) showed a reduction in fatigue at 3 months with ⊕⊕⊝⊝
in short- vention large effect size as measured by FIS subscales (ES (95% CI): Cognitive 0.58 (0.48 to low4
term im- group = 94) 0.68); Physical 0.68 (0.55 to 0.82); Social 0.65 (0.53 to 0.77) and FSS scores: -0.38
pairments (1 study) (-0.45 to -0.31))
3 months
or less af-
ter the start
of the inter-
vention
Measures:
FIS
Follow-up:
up to 3
months

Informed decisions.

Change 299 (inter- One study (Egner 2003, N = 27) showed a reduction of fatigue measured by FSS in ⊕⊕⊝⊝
in long- vention those using video telerehabilitation compared with those using telephone telere- low5
term im- group = 155) habilitation or standard care groups at 6 months (P < 0.05; telephone: SE = 0.478;
pairments (3 studies) standard care: SE = 0.536) and 18 months (P < 0.05; telephone: SE = 0.569; standard
more than 3 care: SE = 0.624). At 12 months follow-up, there was a significant difference in fa-
months af- tigue scores between the video and standard care groups (P < 0.05; SE = 0.471)
ter the in-
tervention One study with 2 reports (Pilutti 2014, N = 82) showed a significant and positive ef-
fect of the intervention on fatigue severity (FSS, P = 0.001) and its physical impact
Measures: (FIS, P = 0.008) at 6-month post-intervention. The results also indicated a favourable
FIS, FSS effect of the intervention on symptoms of pain (MPQ, P =. 0.08) and sleep quality
post-trial (PSQI, P = 0.06), although the differences between groups did not reach
Follow-up: 6 statistical significance
months – 2
years One study (Finlayson 2011, N = 190) showed reduction in fatigue at 6 months with
a large effect size as measured by FIS subscales (ES (95% CI): Cognitive 0.55 (0.46
to 0.64); Physical 0.61 (0.50 to 0.72); Social 0.67 (0.58 to 0.76) and FSS score:: -0.33
(-0.36 to -0.30))
Change in participation
Change in 139 (inter- One study (Egner 2003, N = 27) showed no significant difference in depressive symp- ⊕⊕⊝⊝
psycholog- vention toms measured by CES-D at end of the intervention period (9 weeks). Mean depres- low6
ical out- group = 76) sion scores were lower in those receiving telerehabilitation by video compared with
comes telephone and standard care group symptoms decreased at 6, 8 and 24 months fol-
(3 studies) low-up. Being male was a significant predictor for an increased depression score at
Mea- every measurement point except at 24 months (P < 0.05). Mean CES-D scores fluctu-
sures:CES-D, ated throughout each measurement point for all groups, but seemed to decrease at
HADS, SDMT 24 months in all 3 groups, but not statistically significant. Mean depression scores
were lower in those receiving telerehabilitation by video compared to telephone
Follow-up:
and standard care groups and depressive symptoms also decreased at the 6-, 8- and
variable
24-month follow-ups, but this was not significantly different between groups.
(range 1
month – 2 One study (Paul 2014, N = 30) reported a small non-significant improvement in anx-
years) iety measured by HADS in the control group compared with the treatment group at
post-treatment (8 - 9 weeks) (P = 0.016)

One study with two reports (Pilutti 2014, N = 82) showed a statistically significant
group interaction in psychological outcomes on SDMT scores (F = 5.68, P = 0.02),
which was moderate in magnitude (partial eta squared (ɳ2) = 0.08). There was a clin-
ically meaningful improvement in SDMT scores in the subgroup with mild disability
in the intervention condition (∼ 6 points increase, moderate effect size (d) = 0.41),
whereas those with moderate disability in the intervention condition demonstrat-
ed minimal change (∼ 1 point decrease, d = 0.12). There were minimal changes in
SDMT scores for those with both mild or moderate disability (∼ 1 point increase, d
= 0.10 for both) in the control group. There was also significant improvement in de-
pression and anxiety in the intervention group (with large effect size (ɳ2 = 0.10 for
both) compared with the control group measured by the HADS (depression: F =7.90,
P = 0.006; anxiety: F = 8.00, P = 0.006)
Change in 392 (inter- One study (Egner 2003, N = 27) reported no significant difference in QoL measured ⊕⊕⊝⊝
quality of vention using QWB at the end of the intervention period (9 weeks). Mean QWB scores for low7
life group = 201) each measurement point (6, 9, 12, 18 and 24 months) were higher (indicating higher
QoL) for those in the video group than for the standard care and telephone groups,
Measures: (6 studies, but were significantly better in the video group compared to the telephone group at
QWB, 1 with 2 re- month 12 only (P < 0.05; SE = 0.023). The telephone group and standard care groups
HAQUAMS, ports) reported similar mean QWB scores over the 2-year follow-up period.
MSIS-29,
SF-36, One study (Frevel 2014, N = 18) showed significant improvement in QoL measured
LMSQOLS, by HAQUAMS (cognition: P = 0.026; function of lower limb: P = 0.008; mood: P =
Follow-up: 0.045) in the control group (hippotherapy), but not in the intervention group (e-
variable training)

Informed decisions.

(range 1 One study (Dlugonski 2012, N = 45) showed non-significant condition-by-time inter-
month – 2 actions for QoL measured by MSIS-29. There was no significant correlation between
years) changes in QoL from base line to post-intervention in either the treatment or control
groups
One study (Finlayson 2011, N = 190) showed that significant improvement in HRQoL
in the intervention group on the SF-36 subscales except the physical functioning
and bodily pain subscales: change score (95% CI): Vitality 6.99 (4.29 to 9.69); Role
Emotion 10.08 (4.13 to 16.04); Mental Health 5.78 (3.89 to 7.67); Social Function 7.95
(4.09 to 11.82); General Health 3.61 (1.37 to 5.85); Role Physical 11.12 (6.22 to 16.02)
One study (Paul 2014, N = 30) reported non-significant improvement in HRQoL mea-

sured by LMSQOLS in the treatment group compared with control group post-treat-
ment (8 - 9 weeks) (mean difference -0.07 vs 1.0)
One study with 2 reports (Pilutti 2014, N = 82) reported that participants in the inter-
vention group perceived a positive change in physical HRQoL measured by MSIS-29
(P = 0.06)
Change in other outcomes
Cost effec- 531 (inter- Not measured in any of the studies See 'Impact'
tiveness vention
group = 277)
(9 studies)
Process 80 (interven- One study (Dlugonski 2012, N = 45) showed that participants were most satisfied ⊕⊝⊝⊝
evaluation tion group with (mean ± SD): the overall programme: 4.8 ± 0.4, staff: 4.9 ± 0.2 and pedometer: very low8
(user satis- =46) 4.7 ± 0.6, but slightly less satisfied with the website itself: 4.1 ± 0.9
faction)
(2 studies) One study (Huijgen 2008, N = 35) reported that overall, both participants and thera-
Measures: pists were satisfied with the intervention (over 55% in all 6 items). Both participants
Self-de- and therapists were less satisfied with the aesthetic aspect of the system and had
signed Lik- difficulty completing tasks
ert scale, VAS
scale
Follow-up:
variable
(range 1 - 3
months)
Serious ad- 531 (inter- No serious adverse events reported See 'Impact'
verse events vention
group = 277)
(9 studies)
Care- 531 (inter- Not measured in any of the studies See 'Impact'
givers-re- vention
lated out- group = 277)
comes (9 studies)
ARAT: Action Research Arm Test; CES: Composite Equilibrium Score; CES-D: Center for Epidemiologic Studies Depression Scale; CI:
Confidence interval;DGI: Dynamic Gait Index; EDSS: Expanded Disability Status Scale; ES: Effect size; FIS: Fatigue Impact Scale; FSS:
Fatigue Severity Score; GLTEQ: Godin Leisure-Time Exercise Questionnaire; HADS: Hospital Anxiety and Depression Scale; HAQUAMS:
Hamburg QoL Questionnaire in MS; HRQoL: Health related quality of life; IQR: inter quartile range; LMSQOLS: Leeds MS Quality of
Life Scale; MPQ: McGill Pain Questionnaire; MS: Multiple Sclerosis;MSIS-29: MS Impact Scale; NHPT: Nine Hole Peg Test; PSQI: Pitts-
burg Sleep Quality Index;QoL: quality of life; QWB: Quality of Well- Being Scale; SD: Standard deviation; SDMT: Symbol Disit Modali-
ties Test; SE: Standard Error; SF-36: 36-Item Short Form Health Survey; SOT: Sensory organisation Test; VPR: Visual Preference Ratio;
6MWT: 6 Meters Waltk Test;25FWT: 25 Feet Walk Test; 95% CI: 95 percent confidence interval
# GRADE Working Group grades of evidence

High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Informed decisions.

Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change
the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to
change the estimate.
Very low quality: We are very uncertain about the estimate.
1Methods of randomisation not described or poorly described in 4 studies, only 1 study reported blinding of the assessor, and allocation
concealment was described in only 1 study
2Unclear randomisation procedure, allocation concealment not reported, no blinding of the participants or assessors
3Methods of randomisation not described or poorly described in 1 study, none of the studies reported blinding of the participants or
assessor, and allocation concealment was not described or unclear in 2 studies
4No blinding of the participants or assessors, high risk of attrition bias (> 20% drop-out)
5Methods of randomisation and allocation concealment not described or poorly described in 2 studies, all 3 studies did not report blinding
of the participants or assessor
6Methods of randomisation not described or poorly described in 2 studies, none of the studies reported blinding of the participants or
assessor and allocation concealment procedure
7Methods of randomisation not described or poorly described in 3 studies, allocation concealment procedure described only in 2 studies,
and none of the studies reported blinding of the participants or assessor
8Methods of randomisation and allocation concealment procedure not described or poorly described, and blinding of the participants or
assessor not reported in both studies


Informed decisions.

BACKGROUND method to deliver rehabilitation treatment in a setting convenient

to the patient, such as their home.
Description of the condition
Description of the intervention
Multiple sclerosis (MS) is a chronic neurological disease,
characterised by patchy inflammation, gliosis and demyelination The terminology used in ICT in health care is often
within the central nervous system (CNS), that affects approximately used interchangeably and includes: ‘telemedicine’, ‘telehealth’,
1.3 million people worldwide (WHO 2008). The median estimated ‘telehealthcare’, ‘e-Health’, ‘e-medicine’, ‘telerehabilitation’ etc.
incidence of MS globally is 2.5 per 100,000 (with a range of 1.1 to (Currell 2000; McLean 2010; McLean 2011; Winters 2002). In this
4) (WHO 2008), the prevalence is about 30 per 100,000 population review we define the term ‘telerehabilitation’ as ‘the use of
(range 5 to 80), with a female preponderance (female to male ratio information and communication technologies as a medium for
of 3:1) (Trisolini 2010; WHO 2008). the provision of rehabilitation services to sites or patients that
are at a distance from the provider' (Rogante 2010; Theodoros
The patterns of presentation in MS are heterogeneous and include: 2008). The applications to date encompass systems ranging from
‘relapsing remitting’ (RR) MS (85%), characterised by exacerbations low-bandwidth, low-cost videophones to highly expensive, fully
and remissions; ‘secondary progressive’ (SP) MS with progressive immersive virtual reality systems with haptic interfaces (Theodoros
disability acquired between attacks (in 70% to 75% who start 2008).
with RR, it is estimated more than 50% will develop SPMS
within 10 years, and 90% within 25 years); ‘primary progressive’ Telerehabilitation extends rehabilitative care beyond the hospital
(PP) MS (10%), where persons develop progressive disability process and facilitates multifaceted, often psychotherapeutic
from the onset; and ‘progressive relapsing’ (PR) MS (5%), where approaches to modern management of pwMS at home or in the
persons begin worsening gradually and subsequently start to community (Huijgen 2008). It provides equal access to individuals
experience discrete attacks (MS Australia 2012; Weinshenker 1989). who are geographically remote and to those who are physically
The prognosis in MS is variable and difficult to predict, and and economically disadvantaged (Hailey 2011; Rogante 2010) and
depends on the type, severity and location of demyelinating can improve the quality of rehabilitation delivered (Hailey 2011;
lesions within the CNS (Hammond 2000; MS Australia 2012). Various Kairy 2009; McCue 2010; Rogante 2010; Steel 2011). It can give
factors such as older age at onset, progressive disease course, healthcare providers an opportunity to evaluate the intervention
multiple onset symptoms, pyramidal or cerebellar symptoms and previously prescribed, monitor adverse events and identify areas
a short interval between onset and first relapse are associated in need of improvement. The treating therapists can monitor
with worse prognosis (Hammond 2000). Persons with MS (pwMS) patients’ progress and optimise the timing, intensity and duration
have a prolonged median survival time from the time of of therapy as required, which may not always be possible within the
diagnosis of approximately 40 years (Weinshenker 1989). Therefore, constraints of face-to-face treatment protocols in the current health
issues related to progressive disability (physical and cognitive), systems (Hailey 2011; Steel 2011).
psychosocial adjustment and social re-integration progress over
time. These have implications for pwMS, their carers, treating How the intervention might work
clinicians and society as a whole, in terms of healthcare access, Telerehabilitation is an emerging method of delivering
provision of services and financial burden (Pfleger 2010; Trisolini rehabilitation that uses technology to serve patients, clinicians
2010). and systems by minimising the barriers of distance, time and
cost. The driving force behind this has been the need for an
The pwMS can present with various combinations of deficits
alternative to face-to-face intervention, enabling service delivery in
such as physical (motor weakness, spasticity, sensory dysfunction,
the natural environment – that is, in patients’ homes (Hailey 2011).
visual loss, ataxia), fatigue, pain (neurogenic, musculoskeletal
This method of in vivo delivery of healthcare services can address
and mixed patterns), incontinence (urinary urgency, frequency),
associated issues of efficacy, problems of generalisation and
cognitive (memory, attention), psychosocial, behavioural and
increasing patient participation and satisfaction with treatment.
environmental problems, which limit a person’s activity (function)
and participation (Khan 2007). Cognitive and behavioural problems The benefits and advantages of telerehabilitation have been
can be subtle and often precede physical disability requiring well documented (Bendixen 2009; Brennan 2009; Chumbler
long-term care (Beer 2012). The care needs in this population 2012; Constantinescu 2010; Johansson 2011; Kairy 2009; Lai
are complex due to cumulative effects of the impairments 2004; Legg 2004; Russell 2011; Steel 2011). A home-based
and disabilities, the ‘wear and tear’ and the impact of aging physical telerehabilitation programme was considered feasible
with a disability. Longer-term multidisciplinary management is and effective in improving function in pwMS (Finkelstein 2008).
recommended, both in hospital and in community settings to Telemedicine in pwMS as a tool has the potential for improved
maintain functional gains and social re-integration (participation) health care with reduction in care costs (Zissman 2012). A
over time (Khan 2007; Khan 2010a; WHO 2008). Despite recent systematic review that analysed rehabilitation therapies delivered
advances in MS management, many pwMS are unable to access at home in stroke survivors showed positive outcomes, with
these developments due to limited mobility, fatigue and related a reduction in the risk of deterioration, improved ability to
issues, plus costs associated with travel. With increasing financial perform activities of daily living, reduced costs and duration of
constraints on healthcare systems, alternative methods of service rehabilitation in a frail elderly population (Legg 2004). Other
delivery in the community and over a longer term are now a priority. reports used telerehabilitation to direct multidisciplinary co-
Telerehabilitation for pwMS has potential as a tool to improve ordinated, goal-directed treatment to monitor clinical progress
health care with reduction in care costs (Zissman 2012). The for patients at a distance (Hailey 2011; Kairy 2009; McCue 2010;
emerging advances in information and communication technology Rogante 2010; Steel 2011). In these cases, telerehabilitation
(ICT) may present as an alternative efficient and cost-effective
Informed decisions.

offered an opportunity to provide an individualised rehabilitation Types of interventions

intervention beyond the hospital setting, by regular monitoring
We considered all modalities (type, duration, frequency
and evaluation of the patients' needs and progress, with a
and intensity) of telerehabilitation intervention, using
range of services suited to the individual and their environment
telecommunication technology as the delivery medium, such
(Hailey 2011; Kairy 2009; McCue 2010; Rogante 2010; Steel 2011).
as internet, videoconferencing, telephone and virtual reality,
Telerehabilitation also provides health outcomes comparable
aimed at achieving patient-centred goals or enhancing function
to traditional in-person patient encounters, including improved
and participation. These included: (a) individual (unidisciplinary)
patient satisfaction (Egner 2003; Finkelstein 2008; Hailey 2011;
treatments, e.g. physical interventions: exercise, self-management
Huijgen 2008; Kairy 2009). It can encompass single or multiple
education, etc., and (b) multidisciplinary rehabilitation, i.e.
interventions, or both, aimed at improving the patient experience
delivered by two or more disciplines: occupational therapy,
at the level of impairment, activity or participation, and can
physiotherapy, exercise physiology, orthotics, other allied health
educate patients (and carers) in their ongoing self management.
and nursing, in conjunction with medical input.
Why it is important to do this review The settings of telerehabilitation intervention included the
There is strong evidence to support the effectiveness of following:
rehabilitation programmes for pwMS (Khan 2007; Khan 2010a).
With increasing financial constraints on healthcare systems, • outpatient or day treatment settings in community
alternative methods of service delivery in the community and over rehabilitation centres;
a longer term are now a priority. Telerehabilitation was reported to • home-based settings, in the patients' own homes and local
be effective in various neurological conditions including MS (Egner community.
2003; Finkelstein 2008; Huijgen 2008). However, there is as yet
no systematic review of telerehabilitation interventions in pwMS Control conditions included the following:
to guide treating clinicians on evidence for its validity, reliability,
• no treatment;
effectiveness and efficiency in this population.
• placebo/sham;
This review analyses published and unpublished clinical trials • any type of traditional face-to face rehabilitation treatment in
relating to MS and telerehabilitation, identifies the evidence base outpatient or day treatment settings.
for its use, and discusses issues for future expansion of the evidence
base by traditional research and other methods. We excluded studies if they investigated:
OBJECTIVES • acute medical/surgical/pharmacological interventions for

pwMS provided via telemedicine technology in isolation, unless
To investigate the effectiveness and safety of telerehabilitation it was administered as a concomitant intervention along with
intervention in persons with multiple sclerosis (pwMS) for the telerehabilitation intervention, which was administered in
improved patient outcomes. the same way in both control and treatment groups;
• studies on telerehabilitation targeting mental health conditions
Specifically, the review addresses the following questions: or substance abuse;
• Does telerehabilitation achieve better outcomes compared with • studies on home care (or tele-home care) with no rehabilitation
traditional face-to-face intervention? objectives;
• What types of telerehabilitation interventions are effective, • studies on satisfaction with or acceptance of telerehabilitation
in which setting and influence which specific outcomes technology;
(impairment, activity limitation and participation)? • studies on technical development or feasibility of
telerehabilitation;
METHODS • studies exploring telerehabilitation technology for intra-
professional communication (such as for second opinions) and
Criteria for considering studies for this review for passive information provision, e.g. online education, where
Types of studies there is no direct interaction or involvement of a healthcare
professional with the patient.
We included all randomised controlled trials (RCTs) and controlled
clinical trials (CCTs), including quasi-randomised and quasi- Types of outcome measures
experimental designs with comparative controls (where the
We identified diverse outcomes, given the varied presentations
method of allocation is known but is not considered strictly
of MS-related problems and goals of treatment related to MS
random).
severity. The specific outcome measures per se were not part of
Types of participants the exclusion criteria for this review. We report and list all outcome
measures used in studies in Table 1.
We included studies in pwMS (18 years and over) with a confirmed
diagnosis of MS (Mc Donald 2001; Polman 2005; Poser 1983) and all Primary outcomes
disease subgroups (relapsing remitting, secondary progressive and
We categorised primary outcomes according to the International
progressive MS).
Classification of Functioning, Disability and Health (ICF; WHO 2001),
and included:

Informed decisions.

• improvement in functional activity; such as activities of daily The keywords used to search for studies for this review are listed in
living (ADL), mobility, continence, etc.; Appendix 1.
• improvement in symptoms or impairments, e.g. pain, spasm
frequency, joint range of movement, involuntary movements, Information on the Trial Register of the Review Group and details
spasticity, etc.; of search strategies used to identify trials can be found in
the 'Specialised Register' section within the Cochrane Multiple
• improvement in participation and environmental or personal Sclerosis and Rare Diseases of the Central Nervous System Group
context, or both; e.g. quality of life (QoL), psychosocial function,
module.
employment, education, social and vocational activities, patient
and carer mood, relationships, social integration, etc. Searching other resources
We included the measure of achievement of intended goals for We performed an expanded search to identify articles potentially
treatment, e.g. goal attainment scaling or other measure of goal missed through the database searches and articles from ‘grey
achievement. literature' from 1996 to latest date. This included the following:
It should be noted, however, that some outcome scales crossed • handsearches of reference lists of all retrieved articles, texts and
boundaries between these ICF concepts, for example, items other reviews on the topic;
relating both to impairment (symptoms) and activity. • handsearches of the most relevant journals related to MS and
spasticity research and treatment (such as, but not limited
Secondary outcomes to: Archives of Physical Medicine and Rehabilitation, Journal
These reflect compliance with the intervention, service utilisation, of Rehabilitation Medicine, Journal of Neurology, Journal of
and cost effectiveness of telerehabilitation compared with Neurology, Neurosurgery and Psychiatry, Clinical Rehabilitation,
traditional rehabilitation interventions. Neurology, Physical Therapy, Multiple Sclerosis, Telemedicine
Journal and e-Health, Journal of Medical Internet Research and
We report all adverse events that may have resulted from the others);
intervention. A serious adverse event is defined 'as an event that is • searches using the 'Related articles' feature (via PubMed);
life-threatening or requires prolonged hospitalisation' (Khan 2007). • searches of ProQuest Dissertations and Theses;
We also explored carer-related issues, such as carer strain.
• searches of Web of Science for citation of key authors;
Timing of outcome measures • searches of System for Information on Grey Literature in Europe
The time points for outcome assessments were: short-term (SIGLE);
(immediately after intervention or up to three months) and long- • contacting local and foreign experts for further information,
term (greater than three months) from the start of the intervention. such as MS Groups/Associations, the Cochrane MS Group, key
We considered patient follow-up assessments similarly as short- authors of publications in this review;
term (up to three months) and long-term follow-up (greater than • contacting authors and researchers active in this field.
three months) after cessation of the intervention.
We also searched the following websites for ongoing and
Search methods for identification of studies unpublished trials:
We considered articles in all languages with a view to translation,
• Current Controlled Trials (www.controlled-trials.com);
if necessary. We extracted trials coded with the specific key words
and considered them for inclusion in the review. • UK Clinical Research Network Portfolio Database
(public.ukcrn.org.uk/search/).
Electronic searches
Data collection and analysis
The review authors, along with the Trials Search Co-ordinator,
searched the Cochrane Multiple Sclerosis and Rare Diseases of the Selection of studies
Central Nervous System Group Specialised Register, last searched Two review authors (BA, FK) independently screened and short-
on 9 July 2014, which contains the following: listed all abstracts and titles of studies identified by the search
strategy for appropriateness based on the selection criteria.
1. The Cochrane Central Register of Controlled Trials (CENTRAL)
We independently evaluated each study from the shortlist of
(2014 Issue 7).
potentially appropriate studies for inclusion or exclusion. We
2. MEDLINE (PubMed) (1966 to 9 July 2014). obtained the full text of the article for further assessment to
3. EMBASE (EMBASE.com) (1974 to 9 July 2014). determine if the trial met the inclusion criteria. If we could not
4. Cumulative Index to Nursing and Allied Health Literature reach a consensus about the inclusion or exclusion of any individual
(CINAHL) (EBSCO host) (1981 to 9 July 2014). study, we made a final consensual decision by discussion amongst
5. Latin American and Caribbean Health Science Information all the review authors. We had intended to submit the full article
Database (LILACS) (Bireme) (1982 to 9 July 2014). to the editorial board for arbitration when there was no consensus
6. Clinical trial registries; clinicaltrials.gov. regarding the inclusion or exclusion of a study between the review
authors; however, this was not necessary. We were not masked
7. World Health Organization (WHO) International Clinical Trials
to the name(s) of the study author(s), institution(s) or publication
Registry Portal (apps.who.int/trialsearch/).
source at any level of the review.

Informed decisions.

We had planned to seek further information, where necessary, Measures of treatment effect
about the method of randomisation or a complete description of
A quantitative analysis was not possible due to clinical
the telerehabilitation interventions from the trialists, but this was
heterogeneity (see below), the use of diverse methodology,
not required.
interventions and outcome measures, and insufficient data
Data extraction and management available. We entered and analysed all data in Review Manager 5
software (Review Manager 2014). We qualitatively summarised the
Two review authors (BA, FK) independently extracted data from studies in the Characteristics of included studies tables, presented
each study that met the inclusion criteria, using a standardised data the results of primary and secondary outcomes of included studies,
collection form, with other review authors (JK, MG) making a final categorised according to the ICF framework, in the Summary of
check. We had intended to contact the primary authors of eligible findings for the main comparison. We describe the results in a
studies to provide data and clarification where adequate data were narrative form in the Discussion section below. If studies had been
not reported, but this was not required. We summarise all studies available, and if meta analyses become feasible in future updates,
that met the inclusion criteria in the 'Characteristics of included we will analyse treatment effects as described in the protocol
studies' table provided in Review Manager 5 software developed version of this review (Khan 2013).
by Cochrane (Review Manager 2014), and include details on design,
participants, interventions and outcomes. Unit of analysis issues
We report the following information from individual studies: For each study, we assessed the appropriate units of analysis,
which included the level at which randomisation occurred (e.g.
• publication details; parallel-group design, cluster-randomised trials, cross-over trials,
• study design, study setting, inclusion and exclusion criteria, etc.), type, duration, intensity and setting of telerehabilitation
method of allocation, risk of bias; interventions.
• participant population, e.g. age, type of MS, disease duration, Dealing with missing data
disability (according to Kurtzke's Expanded Disability Status
Scale (EDSS) score (Kurtzke 1982); We provide information about missing data related to participants
• details of intervention; dropping out or lost to follow-up in the Characteristics of
included studies tables. We contacted the primary authors
• outcome measures (primary and secondary);
to obtain additional information and clarification by personal
• withdrawals, compliance, length and method of follow-up and communication (email), to clarify possible overlapping of the data
number of participants followed up. in the four eligible studies. We did not perform imputation of
missing data as we were not able to perform meta-analyses.
We extracted data for every participant assessed for each outcome
measure, and for dichotomous data the number in each treatment Assessment of heterogeneity
group and the numbers experiencing the outcome of interest where
possible. We extracted data for intention-to-treat (ITT) analysis We assessed clinical heterogeneity by examining the characteristics
from each study, and where ITT data were not available, we of studies, the similarity between the types of participants, settings,
retrieved 'on-treatment’ data or the data of those who completed interventions (frequency, intensity, duration) and outcomes, as
the trial. We resolved any disagreement by recourse to other specified in the Criteria for considering studies for this review
review authors (JK, MG) and through discussion, with reference section. Due to apparent clinical heterogeneity, a comprehensive
to the original report. We had planned to contact study authors quantitative analysis (meta-analysis) was not possible. We
for additional information and data if necessary, but this was did not assess statistical heterogeneity and presented the
not required. We present the results in a tabulated format in the studies separately. We will consider both clinical and statistical
Summary of findings for the main comparison. heterogeneity, if data become available in future updates, as
described in the protocol version of this review (Khan 2013).
Assessment of risk of bias in included studies
Assessment of reporting biases
Three review authors (BA, FK, MG) independently assessed the
methodological quality of the included studies using the Cochrane We used a comprehensive search strategy, which included
'Risk of bias' tool (Higgins 2011) for sequence generation, allocation searching for unpublished studies (grey literature), and searching
concealment, blinding of participants, therapists and outcome trials registers (See Search methods for identification of studies) to
assessors, incomplete outcome data and selective outcome avoid reporting biases and publication bias (Egger 1998). We did not
reporting. Further, we also checked baseline data amongst the analyse trial data using funnel plots to investigate the likelihood of
study groups for stability. publication bias, due to the small number of included studies.
We considered a study to be of 'high' methodological quality if the Data synthesis

risks of bias for all domains were low. We termed this a 'high-quality There was a wide variation in several variables of the included
study'. We rated a study as being of 'low' methodological quality studies, such as MS course and severity, content; frequency,
where there was a lack of clarity or a high risk of bias for one or duration, mode of delivery and aim of the interventions; outcome
more domains, and termed this as a 'low-quality study'. If we rated measures used; presentation of results; and methodological
most domains at high risk of bias, we rated the study as a 'very quality. Because of the observed heterogeneity, we did not pool
low-quality study'. We resolved any disagreements by consensus data for a quantitative analysis. If studies had been available and
between the review authors. We present results using 'Risk of bias' if data become available in future updates, we will attempt a
summary figures. quantitative analysis, as described in the protocol version of this
Informed decisions.

review (Khan 2013). We have highlighted the strength of study 4. Cost effectiveness
findings, discussed gaps in the current literature and identified 5. Process evaluation
future research directions in the Discussion section. 6. Serious adverse events
Subgroup analysis and investigation of heterogeneity 7. Caregivers'-related outcomes
We were unable to perform subgroup analysis for the following We used the five GRADE considerations (risk of bias, inconsistency,
subgroups, owing to the lack of available data: imprecision, indirectness and publication bias) to assess the
quality of a body of evidence as it relates to the studies that
1. Type of telerehabilitation intervention (unidisciplinary or contribute data to the meta-analyses for prespecified outcomes.
multidisciplinary, or both). We used the methods and recommendations described in Section
2. Type of MS (relapsing remitting, progressive) 8.5 and Chapter 12 of the Cochrane Handbook for Systematic
3. Severity of MS (i.e. EDSS < 6; > 6) Reviews of Interventions (Higgins 2011) using GRADEpro software
4. Duration of follow-up of participants (≤ 3 months; > 3 months) (GRADEpro 2008). We justified all decisions to downgrade or
upgrade the quality of studies by using footnotes, and we made
Sensitivity analysis comments to aid readers' understanding of the review when
necessary.
We were not able to conduct sensitivity analyses due to our
narrative presentation of the results of the included studies. If RESULTS
studies had been available, and heterogeneity existed across trials,
we would have conducted sensitivity analyses by omitting trials Description of studies
with a high risk of bias as described in the Cochrane Handbook for
Systematic Reviews of Interventions (Higgins 2011). If meta-analyses See: Characteristics of included studies; Characteristics of excluded
become feasible in future updates, we will perform sensitivity studies
analyses as described in the protocol version of this review (Khan
Results of the search
2013).
Electronic and manual searches identified 4030 references
'Summary of findings' table (MEDLINE = 79; EMBASE = 3799; CENTRAL = 136; CINAHL = 5;
LILACS = 9; CRD database = 0; Cochrane Opportunity Fund Project
These outcomes are included in the Summary of findings for the
= 0; Trial Registries via WHO Portal = 0; handsearching journals
main comparison:
= 0; handsearching trial registries = 2) with our search criteria.
1. Change in disability (post-intervention, ≤ 3 months, > 3 months) After elimination of duplicates records, we screened the remaining
3842 for closer scrutiny. Of these, we retrieved the full text of 29
2. Change in impairments (post-intervention, ≤ 3 months, > 3
articles for further assessment to determine inclusion in the review.
months)
We did not identify any ongoing or unpublished studies awaiting
3. Change in participation (psychological outcomes, QoL) classification. See: Figure 1 for Study flow chart.


Informed decisions.

Figure 1. Study flow diagram.

Included studies Motl 2011; Paul 2014; Pilutti 2014) fulfilled the inclusion criteria for
this review (see Characteristics of included studies table).
In total, nine RCTs, one with two reports (Pilutti 2014; Sandroff
2014), published between 2003 and 2014 (Dlugonski 2012; Egner Five of the included studies were conducted in the United States
2003; Finlayson 2011; Frevel 2014; Gutíerrez 2013a; Huijgen 2008; (Dlugonski 2012; Egner 2003; Finlayson 2011; Motl 2011; Pilutti

Informed decisions.

2014); one each was conducted in Spain (Gutíerrez 2013a), Germany the effectiveness of an individualised virtual reality
(Frevel 2014) and the United Kingdom (Paul 2014), while one was telerehabilitation programme for improvement in postural
a multicentre study conducted in three different countries (Italy, control
Spain and Belgium; Huijgen 2008). Three studies were conducted • One study examined the effectiveness of an internet-based
by the same group of authors in the same setting and with the same home training programme (e-Training) in comparison with
cohort of participants recruited from a single database (Dlugonski hippotherapy to improve balance (Frevel 2014)
2012; Motl 2011; Pilutti 2014), of which one reported different
outcomes in two different articles (Pilutti 2014). The duration and intensity of the telerehabilitation interventions
varied significantly depending on the nature of the intervention,
Participants and ranged from one to six months (median 12 weeks). None
Participants' detailed information, including inclusion/exclusion of the studies reported the recruitment time period. The follow-
criteria and baseline demographics, are listed in the Characteristics up periods varied between trials, but all studies assessed the
of included studies table. The nine included studies involved a participants immediately after intervention. Only one trial reported
total of 531 participants (277 participants in the treatment groups long-term follow-up of up to 24 months (Egner 2003). For details of
and 254 in the control groups). The number of participants in the assessment time points for each trial refer to the Characteristics of
studies ranged from 27 to 190 (median 45). As expected, there included studies tables.
were more women, with their proportion ranging from 56% to 87%
Excluded studies
(mean 74%). The mean age of participants varied from 41 to 52
years (mean 46.5 years) and mean years since diagnosis from 7.7 We excluded 16 studies after appraisals of the full reports (listed in
to 19.0 years (mean 12.3 years). The majority of participants had the Characteristics of excluded studies tables). The primary reason
a relapsing-remitting course of MS (RRMS), two studies involved for exclusion was:
only people with RRMS (Dlugonski 2012; Motl 2011) and two studies
did not provide details of MS type (Egner 2003; Huijgen 2008). The • 10 studies addressed mental health care as a primary
study inclusion criteria varied between trials. All trials included intervention (Amato 2014; Beckner 2010; Cerasa 2013; Fischer
participants with definite MS, although only two trials specified the 2013; Mohr 2000; Mohr 2005; Mohr 2007; Moss-Morris 2012;
commonly-used McDonald's criteria (Mc Donald 2001) (Frevel 2014; Solari 2004; Stuifbergen 2012)
Gutíerrez 2013a). One study reported secondary data regarding • One study had a medical-care intervention only (Zissman 2012)
MS participants which were collected as part of a larger study of • One study evaluated the effectiveness of an online fatigue
a telerehabilitation intervention in people with severe mobility self-management programme for people with various chronic
impairment (Egner 2003). neurological conditions including MS, but did not provide a
subgroup analysis for the MS cohort (Ghahari 2010)
Intervention
• Two studies assessed counselling interventions for health
Detailed information about interventions in the included studies promotion and major depression (Bombardier 2008;
is presented in the Characteristics of included studies tables and Bombardier 2013)
is further summarised in Table 2. The various telerehabilitation • Two studies assessed interventions with no rehabilitation
interventions in the included studies consisted generally of physical objectives, such as education, self management (Miller 2011;
activity and educational components. Wiles 2003)
• Three studies used similar internet-delivered, social cognitive Risk of bias in included studies
theory-based behavioural intervention to increase physical
activity (Dlugonski 2012; Motl 2011; Pilutti 2014) See: ’Risk of bias’ tables in the Characteristics of included studies
• One study evaluated a structured in-home education and and Figure 2 and Figure 3.
counselling session delivered via telephone or video by a
Figure 2 and Figure 3 represent the review authors’ judgements
rehabilitation nurse (Egner 2003)
about each methodological quality item, presented as percentages
• One study examined a group-based, teleconference-delivered across all included studies and a summary of the risk of
fatigue management programme (Finlayson 2011) bias, respectively. Where studies failed to report sufficient
• One study evaluated a telerehabilitation intervention for arm/ methodological detail to assess the potential risk of bias, we graded
hand function at home - the 'Home Care Activity Desk' (HCAD), them as being at 'unclear’ risk (presented as symbol '?' in Figure
which consists of a set of exercises for functional activity of the 3). The methodological quality of the nine included trials was
upper limb (Huijgen 2008) 'low’, with substantial flaws in the methodological design and a
• One study evaluated the effectiveness of an individualised web- high risk of bias related to their randomisation procedure; blinding
based physiotherapy programme (Paul 2014) of participants, therapists and outcome assessors, and outcome
• One study published in two different journals by the analysis.
same authors (Gutíerrez 2013a; Gutierrez 2013b) examined


Informed decisions.

Figure 2. Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages
across all included studies.


Informed decisions.

Figure 3. Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Informed decisions.

Allocation Furthermore, this series of studies published one trial (Pilutti 2014)
with different outcomes in another report (Sandroff 2014). Most
Although all included studies stated that the procedure was
included studies had short-term follow-up, and were restricted
randomised, the methods of randomisation were adequately
to immediate post-treatment assessments. Most studies seemed
reported in only six studies (one with two reports) (Dlugonski 2012;
to be underpowered and only one study performed a sample
Finlayson 2011; Frevel 2014; Motl 2011; Paul 2014; Pilutti 2014).
size calculation (Finlayson 2011). One study (Egner 2003 ) failed
• Two studies used a random number generator for to report the participant recruitment process and methodology
randomisation (Dlugonski 2012; Pilutti 2014) in detail, and allocation of participants to treatment and control
groups was unbalanced in two studies (Egner 2003; Huijgen 2008).
• One study used a random permutated block design (Finlayson
2011) Effects of interventions
• One randomly allocated the participants using simple allocation
by drawing lots of preshuffled opaque envelopes (Frevel 2014) See: Summary of findings for the main comparison
• One study used a series of random numbers generated in Meta-analysis was not possible due to the heterogeneity of the
Microsoft Excel (consecutive numbers allocated, where even included studies mentioned earlier. The included studies used a
numbers represented the intervention group and odd numbers range of telerehabilitation approaches in pwMS (see Table 2 for
the control group) (Paul 2014) the summary of telerehabilitation interventions) and a broad range
of outcome measures (see Table 3 for a list of outcome measure
Only three studies described in detail concealment of allocation
used). A summary of the findings of the included trials is presented
prior to entry to the study (Finlayson 2011; Frevel 2014; Motl
based on primary and secondary outcomes categorised according
2011). Other studies either gave little or no information about
to the International Classification of Functioning, Disability and
the randomisation procedure, or used non-random components
Health (ICF) framework in the Summary of findings for the
like alternation, assignment to comparable groups with respect to
main comparison. Pooling of data from the included studies was
clinical and demographic factors, or allocation of participants to
confounded by the differences between interventions and the use
the intervention group after initial randomisation.
of different outcome measures, as highlighted above.
Blinding
Primary outcomes
Blinding of participants and treating personnel can be challenging
Improvement in functional activity
in rehabilitation trials, because of the characteristics of
interventions. However, blinding of outcome assessors is possible All studies except two (Egner 2003; Finlayson 2011) assessed the
and highly desirable (Amatya 2013). The blinding of participants first prespecified primary endpoint to improve functional activity
and personnel was insufficiently reported in most of the studies. in pwMS (N = 314 participants,low quality evidence). All studies
Only one study took measures to blind participants to group evaluated participants immediately after the intervention, using
allocation (Finlayson 2011). None of the studies attempted to different instruments (see Table 3 and Summary of findings for the
blind the treating personnel. One study mentioned blinding of the main comparison), with intervention periods ranging from one to
outcome assessors, but provided no details (Gutíerrez 2013a). six months. Overall six studies assessed the functional endpoint
post-intervention up to 12 weeks (Dlugonski 2012; Frevel 2014;
Incomplete outcome data Gutíerrez 2013a; Huijgen 2008; Motl 2011; Paul 2014).
The drop-out rate of participants during the trial period ranged
Two studies (Dlugonski 2012; Motl 2011) conducted in different time
from 0% to 21%. In four studies, there were no or minimal losses
periods with the same cohort of participants showed significant
to follow-up (Dlugonski 2012; Egner 2003; Gutíerrez 2013a; Paul
improvement in physical activity in the treatment group at the post-
2014). Drop-outs and withdrawals were higher than 20% in only
intervention assessment, as measured by the Godin Leisure-Time
one study (Finlayson 2011), which recruited the highest number of
Exercise Questionnaire (GLTEQ) (P < 0.01). The authors' reported
participants. One study which included MS participants as one of
increase in physical activity was sustained at three-month follow-
the subgroups failed to report the attrition rate (Huijgen 2008). Most
up compared with the control group (P < 0.001) (Motl 2011).
of the studies did not conduct intention-to-treat analysis.
One study (Frevel 2014) comparing two interventions, e-training
Selective reporting
and hippotherapy, showed significant improvement in dynamic
All the included studies reported prespecified (primary and and static balance capacity compared with baseline values in both
secondary) outcomes (see Table 1 and Table 3 for a list of the the intervention (e-training) (Dynamic Gait Index (DGI): P = 0.016,
outcome measures). Berg Balance Scale (BBS): P = 0.011) and control (hippotherapy)
groups (DGI: P = 0.011, BBS: P = 0.011). However, there was no
Other potential sources of bias difference between groups.
Sample sizes were small (< 40 participants) in four studies (Egner
Huijgen 2008 showed no statistically significant differences
2003; Frevel 2014; Huijgen 2008; Paul 2014). A series of three studies
between the intervention using telerehabilitation for arm functions
was conducted by the same group of authors, which recruited
(Home Care Activity Desk (HCAD)) and control groups in arm
selective participants who volunteered for research through a
function as measured by Action Research Arm Test (ARAT) (mean
single database for the same institutions (Dlugonski 2012; Motl
change 1.26, 90% confidence interval (CI) -1.90 to 4.42) and Nine-
2011; Pilutti 2014). Although none of these studies mentioned
Hole Peg Test (NHPT) (mean change 7.24, 90% CI -6.55 to 23.25).
overlapping of the recruited participants, we cannot rule out the
possibility of inclusion of the same participants in different trials.
Informed decisions.

Paul 2014 reported an increase in gait speed using the 25 Foot Walk One study (Pilutti 2014) showed a significant positive effect of the
Test (25FWT) in the intervention group compared with the control behavioural intervention on fatigue severity (FSS, P = 0.001) and
group, but this was not statistically significant (P = 0.170). The its physical impact (FIS, P = 0.008) at six-month post-intervention.
intervention group had a statistically significant improvement in There was a favourable effect of the intervention on symptoms
the physical subscale of the Multiple Sclerosis Impact Scale (MSIS) of pain (McGill Pain Questionnaire (MPQ), P = 0.08) and sleep
(P = 0.048). quality post-trial (Pittsburgh Sleep Quality Index (PSQI), P = 0.06),
although the differences between groups did not reach statistical
Another study (Gutíerrez 2013a) showed improvements in balance significance.
and postural control, with a significant increase in Composite
Equilibrium Score (CES) in the intervention group (mean change Frevel 2014 reported significant improvement in fatigue in the
8.21 points, P < 0.001), but not in the control group (mean control group (hippotherapy) (P < 0.05) for all subscales of the
change 1.93, P = 0.123). Visual Preference Ratio and contribution Modified Fatigue Impact Scale (MFIS), while the intervention group
of vestibular information (VEST, Vestibular Ratio) improved (e-training) improved only on the MFIS cognitive subscale (P =
significantly in the intervention group (P < 0.001), but not in the 0.031). A significant difference between the groups was noted only
control group (P > 0.05). There were significant post-treatment in the cognitive subscale of the MFIS ( P = 0.012).
differences between treatment and control groups in the CES (F =
37.873, P < 0.001) and the VEST (F = 12.156, P < 0.001). Significant One study (Paul 2014) reported no improvements in symptoms as
post-treatment differences between groups were also found for measured by the MS Symptom Checklist.
the ability to accept incorrect visual information expressed by the
Improvement in participation
visual conflict parameter (F = 15.05, P < 0.000), which demonstrates
that the treatment group showed a greater ability to accept post- Psychological outcomes
treatment afferent inputs compared with the control groups. There
Overall three studies (one with two reports), assessed cognitive
were no significant between-group differences in the contribution
functions as one of the outcomes (N = 139 participants, low quality
of the visual system (F = 2.64, P = 0.11) or use of somatosensory
evidence) (Egner 2003; Paul 2014; Pilutti 2014).
information (F = 0.117, P = 0.734) in the maintenance of balance and
stability. Egner 2003 showed that a telerehabilitation intervention
(structured in-home counselling and education) delivered via
One study (Sandroff 2014) evaluating an internet-delivered
telephone or video, improved depressive symptoms as measured
behavioural intervention, showed a significant positive effect of
by the Centre for Epidemiologic Studies Depression Scale (CES-D)
the intervention on the Six Minute Walk (6MW) test relative to the
at the end of the intervention period (nine weeks) in both groups.
control group (P = 0.07). The authors also found physical activity
Mean CES-D scores fluctuated, but decreased at 24 months in all
increased most in those with mild disability.
three groups. This was, however, not statistically significant. Mean
Improvement in impairments depression scores were lower in those receiving telerehabilitation
by video compared with telephone and standard-care groups, and
Five studies assessed the prespecified primary endpoint depressive symptoms also decreased at the six-, eight- and 24-
(improvement in impairments) using different measures (N = 347 months follow-ups, but this was not significantly different between
participants;low quality evidence) (Egner 2003; Finlayson 2011; groups. The authors reported that being male was a significant
Frevel 2014; Paul 2014; Pilutti 2014). predictor for increased depression score at every measurement
point except at 24 months (P < 0.05) (Egner 2003).
Fatigue was the primary outcome in three studies (Egner 2003;
Finlayson 2011; Pilutti 2014), all reporting significant differences Paul 2014 reported a small non-significant improvement in anxiety
between groups in favour of the intervention group. One study measured by the Hospital Anxiety and Depression Scale (HADS) in
(Finlayson 2011) showed a significant reduction in fatigue in the the control group compared to the treatment group post-treatment
intervention group immediately after intervention compared to a (eight to nine weeks) (P = 0.016).
wait-list control group as measured by the Fatigue Impact Scale
(FIS) in all three subscales: mean difference (SD): Cognitive -3.12 One study with two reports (Pilutti 2014) showed a statistically
(6.1), P = 0.001; Physical -2.53 (6.4), P = 0.014; Social -6.01 (12.1), significant group interaction in psychological outcomes on Symbol
P = 0.002. These changes were maintained with large effect sizes Digit Modalities Test (SDMT) scores (F = 5.68, P = 0.02), which was
in all FIS subscales at three-month follow-up: Effect Size (95% CI): moderate in magnitude (partial eta squared (ɳ2) = 0.08). There
Cognitive 0.58 (0.48 to 0.68); Physical 0.68 (0.55 to 0.82); Social 0.65 was a clinically meaningful improvement in SDMT scores in the
(0.53 to 0.77), and at six-month follow-up: Cognition: 0.55 (0.46 to subgroup with mild disability in the intervention condition (∼ 6
0.64); Physical: 0.61 (0.5 to 0.72) and Social: (0.67 (0.58 to 0.76). points increase, moderate effect size (d) = 0.41), whereas those with
There was also a significant reduction in the Fatigue Severity Scale moderate disability in the intervention condition demonstrated
(FSS) scores at all three time periods. minimal change (∼ 1 point decrease, d = 0.12). There were minimal
changes in SDMT scores for those with mild or moderate disability
Egner 2003 analysed the impact of a telerehabilitation intervention (∼ 1 point increase, d = 0.10 for both) in the control group.
(structured in-home counselling and education) delivered via There was also significant improvement in depression and anxiety
telephone or video, and reported similar fatigue scores (measured in the intervention group (with large effect size (ɳ2 = 0.10 for
using FSS) for all three groups (video, telephone and standard both) compared with the control group measured by the HADS
care) at nine weeks post-intervention; however, the participants in (depression: F =7.90, P = 0.006; anxiety: F = 8.00, P = 0.006) (Pilutti
the video group had significantly lower scores than the other two 2014).
groups at six months (P < 0.05) and at 18 months (P < 0.05).

Informed decisions.

Quality of life No studies reported data on cost effectiveness, investment costs

Six studies assessed quality of life (QoL), using different outcome or resource utilisation. None of the included studies reported
measures (N = 392 participants;low quality evidence) (Dlugonski any serious adverse effects attributable to telerehabilitation. carer
2012; Egner 2003; Finlayson 2011; Frevel 2014; Paul 2014; Pilutti burden or social integration (in the form of return to work, study
2014). etc.) were not evaluated in any of the studies.
Egner 2003 reported no significant difference in QoL between the DISCUSSION

treatment groups (video or telephone) and control group (standard
This review investigated the effectiveness of different forms
care) measured using the Quality of Well-Being Scale (QWB) at the
of organised telerehabilitation in adults with multiple sclerosis
end of the intervention period (nine weeks). However, mean QWB
(MS) on measures of activities, impairments and participation
scores for each measurement point (6-, 9-, 12-, 18- and 24-months)
based on the International Classification of Functioning, Disability
were higher (indicating higher QoL) for participants in the video
and Health (ICF) framework (WHO 2001), and also of the
group than for those in the standard care and telephone groups.
safety and cost effectiveness of these interventions. There was
There were significantly higher QWB scores in the video compared
marked heterogeneity between the included trials in terms of
with the telephone groups at 12 months follow-up only (P < 0.05;
characteristics, type and mode of delivery of the telerehabilitation
standard error (SE) = 0.023). The telephone group and standard-
interventions, measurement tools used (even for identical
care groups reported similar mean QWB scores over the two-year
outcomes), treatment and control protocols and length of follow-
follow-up period (Egner 2003).
up. We therefore performed a best-evidence synthesis using a
One study (Frevel 2014) showed significant improvement in QoL qualitative analysis.
measured by the Hamburg Quality of Life Questionnaire in Multiple
Sclerosis (HAQUAMS) (in subscales - Cognition: P = 0.026; Function
Summary of main results
of lower limb: P = 0.008; Mood: P = 0.045) in the control group This review of nine randomised controlled trials (RCTs) (one
(hippotherapy), but not in the intervention group (e-training). with two reports), involving 531 participants with MS (N =
277 participants in the intervention group) evaluated a wide
Finlayson 2011 reported that a fatigue management programme variety of telerehabilitation interventions (see Table 2). All
showed significant improvement in QoL in the intervention group telerehabilitation interventions were complex, using more than
on the 36-item Short Form Health Survey Questionnaire (SF-36) in one active rehabilitation component which differed in many
all subscales except physical functioning and bodily pain (change aspects, including intervention goals, number and extent of
score (95% CI)): Vitality 6.99 (4.29 to 9.69); Role Emotion 10.08 (4.13 the intervention components, duration and intensity, and mode
to 16.04); Mental Health 5.78 (3.89 to 7.67); Social Function 7.95 of delivery. Control interventions also differed between studies
(4.09 to 11.82); General Health 3.61 (1.37 to 5.85); Role Physical ranging from 'usual care' or 'wait-list' to active intervention
11.12 (6.22 to 16.02). (such as hippotherapy, Frevel 2014). Most interventions included
physical activity as one of the main intervention components,
Two studies (Dlugonski 2012; Pilutti 2014) assessed QoL using
followed by education and behavioural training. The included trials
the Multiple Sclerosis Impact Scale (MSIS-29) and found no
were heterogeneous in terms of outcome measures used and
significant correlation between changes in QoL from baseline to
study quality. Quantitative synthesis was therefore not possible.
post-intervention in either treatment or control groups.
A qualitative synthesis of 'best evidence’ for telerehabilitation
Similar non-significant improvement in QoL was reported in interventions indicates low level evidence for:
another study (Paul 2014), at post-treatment (eight to nine weeks),
• Short-term benefit in improving functional activities, such
in which authors used the Leeds Multiple Sclerosis Quality of Life
as physical activity, balance capacity and postural control
Scale (LMSQOL) (mean difference in treatment group: -0.07 versus
compared with baseline, and some benefit in improving
control group: 1.0).
walking, physical activity;
Secondary outcomes • Short-term benefit in reducing and/or improving impairments,
such as fatigue, and long-term benefits in improving symptoms
Two studies (Dlugonski 2012; Huijgen 2008) reported process
such as fatigue, pain and insomnia;
evaluation (satisfaction and acceptance of the telerehabilitation).
• Longer-term improvement in participation, such as improving
Dlugonski 2012 used a five-item Likert satisfaction scale and found psychological outcomes and quality of life (QoL)
that participants were most satisfied with the overall programme
(mean ± SD):4.8 ± 0.4, staff: 4.9 ± 0.2 and pedometer: 4.7 ± 0.6, but There is a 'very low' level of evidence for participants' and
slightly less satisfied with the website: 4.1 ± 0.9. therapists' satisfaction with the telerehabilitation interventions.
Huijgen 2008 used a six-item Visual Analogue Scale (VAS) to The quality of evidence is further compromised by the limited
evaluate users' and therapists' satisfaction with the upper limb number of studies, heterogeneity and the methodological
telerehabilitation intervention. Overall, both participants and weaknesses identified (underpowered with small sample sizes,
therapists were satisfied with the intervention (over 55% in all six high risk of bias, short follow-up periods, lack of rigorous
items). The authors found that both participants and therapists methodology and different outcome measures) amongst the
were less satisfied with the aesthetic aspect of the intervention and included trials.
had difficulty in completing prescribed tasks. Subgroup analysis for type of telerehabilitation intervention
(unidisciplinary or multidisciplinary, or both), type of MS (relapsing

Informed decisions.

remitting, progressive), severity of MS (Expanded Disability Status • Lack of longer-term follow-up to detect the long-term effects of
Scale (EDSS) < 6; > 6) and duration of follow-up of participants intervention; only three studies (one with two reports) followed
(≤ 3 months; > 3 months) was not possible due to lack the participants beyond three months (Egner 2003; Finlayson
of data. There were no data for the cost effectiveness of 2011; Pilutti 2014);
telerehabilitation interventions, their impact on health service • Lack of control for participants’ personal and other
utilisation (hospitalisation or attendance/access to the health confounding factors, which influence patient-therapist
services) and carer burden or social integration (in the form of interaction, compliance, and delivery of therapy, thus impacting
return to work, study etc.). There were limited data on process on outcomes such as participant motivation and self efficacy,
evaluation (satisfaction and acceptance of the telerehabilitation) comorbidity and activity level outside of therapy programmes
and no reports of serious adverse effects attributable to (not assessed in any of the studies).
telerehabilitation.
Potential biases in the review process
Overall completeness and applicability of evidence
We conducted the search in conjunction with the Trials Search Co-
Overall, this review indicates that telerehabilitation has some ordinator from the Cochrane Multiple Sclerosis and Rare Diseases
impact on improving function and symptoms (including cognitive of the Central Nervous System Working Group in the Cochrane
function), but does not have an appreciable impact on disease- MS Group Specialised Register using a broad search strategy. In
specific QoL in persons with MS (pwMS). There are no cost order to avoid publication bias, we performed literature searches
data or data on hospitalisation or access to other services. As at three different time points. This process would have captured
aforementioned, there was marked variation between studies both published and ongoing trials coded as MS by the Cochrane
concerning the content and mode of delivery of the interventions. MS Group. Two review authors further selected relevant articles
This highlights the diversity of programmes currently offered to from this extensive list independently and agreed on a final list
pwMS. of included studies by consensus between all four review authors.
We applied no language restriction, although all the included
Pooling data for meta-analyses to make meaningful statements trials were published in English. Overall, the review methodology
for both primary and secondary outcomes was not possible. is comprehensive, as described in the Cochrane Handbook for
The generalisability and applicability of the results are limited, Systematic Reviews of Interventions (Higgins 2011). However, we
as most studies recruited participants from a single centre with recognise a number of limitations in the methodology of the review
strict inclusion and exclusion criteria. Moreover, generalisability of itself, and the completeness of the retrieved literature:
results to different countries and healthcare systems also seems
limited, as the studies were conducted predominantly in the USA 1. Four of the included studies in this review (Dlugonski 2012; Motl
and Europe. 2011; Pilutti 2014; Sandroff 2014) were conducted by the same
group of authors in the same cohort of participants recruited
Quality of the evidence from a single database, and using the same behavioural
In general, we rated the nine included studies (one with two intervention (modified in recent publications). Hence, we
reports) as of 'low' methodological quality due to substantial cannot rule out overlapping of the participants amongst the
flaws in their methodological design with various biases observed. studies.
These included a lack of proper randomisation, problems with 2. We categorised outcomes according to the World Health
allocation concealment and a lack of blinding. Further, there was Organization (WHO) ICF, which might have posed some
also insufficient information about these specific methodological methodological problems, since many of the outcome measures
issues, so that many domains of the 'Risk of bias' tables are rated used in the included trials crossed the boundaries between
as 'unclear’ (see Figure 2 and Figure 3). All studies except one the different levels of the ICF model. However, this model is
were single-centre trials, with fairly small participant numbers, with widely used worldwide and helpful in clarifying the experience
a concomitant risk of type I and II errors. The evidence is very of people who live with long-term neurological conditions, such
heterogeneous, particularly in terms of interventions (technology as MS (Khan 2007).
employed, rehabilitation components within the intervention, 3. We cannot rule out some degree of selection bias from the
duration and intensity of the intervention etc.), and diverse literature search (Van Tulder 2003), given that the search
outcome measures used. The other methodological flaws include: strategy principally encompassed the cited literature, despite
the extended range of terms for both MS and telerehabilitation
• High risk of selection bias, as only three studies (Finlayson 2011; that we used to capture the widest possible selection from the
Frevel 2014; Sandroff 2014) described allocation concealment relevant literature.
• Lack of description of the randomisation procedure, adequately 4. We cannot rule out publication bias as we cannot exclude the
reported in only three studies (Finlayson 2011; Frevel 2014; Paul possibility that there have been negative trials that have not
2014) reached the published literature (Egger 1998).
• High risk of performance bias due to non-blinding of the study 5. Reference bias (Gøetzsche 1987) is possible, as we searched
participants and treating personnel; participants were blinded the bibliography lists of only relevant papers for other possible
to group allocation in only one study (Finlayson 2011), but articles missed in our electronic searches.
treating personnel were not blinded; and only one study took
measures to blind outcome assessors (Gutíerrez 2013a) We therefore welcome contact from any readers who are aware of
• Most studies were underpowered with small sample sizes important studies that would meet the criteria for this review, but
• Lack of an intention-to-treat analysis protocol in most trials have not so far been included.

Informed decisions.

Agreements and disagreements with other studies or Kairy 2009; Rogante 2010). It can give healthcare providers an
reviews opportunity not only to evaluate the interventions previously
prescribed, but also to monitor adverse events and identify areas
To date, there has been no systematic review assessing the in need of improvement by evaluating patients’ progress (McCue
effectiveness of telerehabilitation in pwMS to guide treating 2010). Moreover, it provides an opportunity to optimise the timing,
clinicians or policy makers. Positive effects and successful intensity and duration of therapy as required, which may not
implementation of telerehabilitation were reported in various always be possible within the constraints of face-to-face treatment
neurological conditions including stroke (Johansson 2011; Legg protocols and scheduling in current health systems (Hailey 2011;
2004), Parkinson’s disease (Giansanti 2008) and other non- Steel 2011). MS is a complex and challenging condition requiring
neurological conditions such as musculoskeletal conditions individualised and integrated multidisciplinary care. The range of
(Russell 2011; Tousignant 2011), injuries (Bendixen 2008; Forducey telerehabilitation interventions and their intensity requirements
2003; Houlihan 2011) and chronic diseases (Steel 2011). We can vary from person to person and are difficult to standardise.
found one systematic review (Hailey 2011) (also published Various factors such as the patient's personal characteristics, their
earlier as a health technology assessment, Hailey 2010), with comorbidities, functional and coping abilities, family dynamics,
some overlap with our results. That review considered the and the healthcare system may impact patient outcomes (Khan
evidence of benefit from the use of telerehabilitation for various 2010b). There is a paucity of information on the interaction of these
conditions, including neurological ones. The authors conducted factors on patient outcomes and very little is understood about the
comprehensive searches in multiple databases up to November 'black box' of rehabilitation in the MS population (Khan 2010b)
2009 and included two studies (one observational and one
RCT) on telerehabilitation in the management of people with This review highlights the lack of robust, methodologically-
MS. That review provided simply an overview of studies on strong studies evaluating the effectiveness of telerehabilitation
telerehabilitation for certain groups of conditions in terms of intervention in this population. Overall, the review found low
feasibility of interventions, the clinical significance of results, and quality evidence for a beneficial effect of telerehabilitation
a requirement for further data to establish the application as interventions on reducing short-term disability and impairments,
suitable for routine use. Consistent with the results of our review, such as fatigue. There was also low-quality of evidence suggesting
the authors found inconsistent or insufficient evidence of benefit some benefit in improving functional activities and impairments
for telerehabilitation interventions and their impact on routine in the longer term, and improving psychological outcomes and
rehabilitation programmes. quality of life (QoL). There are limited data on process evaluation
(participants' and therapists' satisfaction) and, surprisingly, none
AUTHORS' CONCLUSIONS of the studies addressed cost effectiveness.
Implications for practice Telerehabilitation has a major role in providing remote

rehabilitation to people with chronic neurological conditions in
Multiple sclerosis (MS) is a complex condition with different
future, and has potential to fill the existing service gap in the
patterns of presentation and variable prognoses. The care needs
care of pwMS. However, the clinical applicability of the findings of
in this population are complex due to cumulative effects of
this review and the effectiveness of telerehabilitation interventions
the impairments and disabilities, the ‘wear and tear’, and the
need to be confirmed in future research.
impact of aging with a disability. Therefore, issues related to
progressive disability, psychosocial adjustment and social re- Implications for research
integration progress over time need to be considered. These
have implications for persons with MS (pwMS), their family/carers, This review found various limitations and gaps in knowledge, which
treating clinicians and society as a whole, in terms of healthcare could suggest directions for future research. These include, but are
access, provision of services and financial burden (Beer 2012; not limited to:
Khan 2010a). Multidisciplinary rehabilitation is recommended
and proven to be effective for pwMS, both in hospital and in • More methodologically robust studies, e.g. randomised
communities, to maintain functional gains and social re-integration controlled trials (RCTs) comparing different models and
(participation) (Khan 2007). However, many pwMS are unable to intensity of telerehabilitation
access appropriate treatment due to limited mobility, fatigue and • Large-scale systematic and 'practice-based trials' in which data
related issues, and limited access to services, and the costs and are routinely gathered without disrupting the natural milieu of
time associated with travel. treatment to provide valuable information about outcomes in
real-life clinical settings
With advances in information and communication technology, new • Use of more sensitive and appropriate validated outcome
models of care such as telerehabilitation can be an alternative measures that are important for patients and their
efficient and cost-effective method to deliver rehabilitation. The MS representatives and that focus on impairments, activity
population is likely to be receptive to and benefit from this type limitations and restriction in participation
of care model as most are young and have high rates of internet
• Longitudinal data in the MS population to ascertain long-term
use (Motl 2011; NMSS 2007). Telerehabilitation is an alternative
care needs
to traditional face-to-face interventions, providing equal access
for individuals who are geographically remote and for those who • More research about patient and carer perspectives and their
are physically and economically disadvantaged, and can improve involvement in telerehabilitation
the quality of rehabilitation delivered by addressing associated • Research about specific telerehabilitation modalities and
issues of efficacy, problems of generalisation and increasing interventions in MS to improve evidence-based practices
patient participation and satisfaction with treatment (Hailey 2011; • Cost effectiveness of telerehabilitation
Informed decisions.

• More emphasis on participatory domains (cognitive outcomes ACKNOWLEDGEMENTS

and quality of life (QoL)) in MS for impact on societal integration
We thank the Editorial Board and Ms Liliana Coco, the Managing
Future studies in telerehabilitation should focus on improving Editor, of the Cochrane Multiple Sclerosis and Rare Diseases of
the methodological and scientific rigour of clinical trials, with the Central Nervous System Review Group for their support and
larger sample sizes and with longer-term follow-up. Further, active assistance. We are grateful to Ms Andrea Fittipaldo, the Trials Search
clinician involvement is needed to build evidence in this area for Co-ordinator, for literature searches, and Dr. Paola Mosconi for
everyday clinical practice. reviewing the manuscript.

Informed decisions.

REFERENCES

References to studies included in this review Sandroff BM, Klaren RE, Pilutti LA, Dlugonski D, Benedict RH,
Motl RW. Randomized controlled trial of physical activity,
Dlugonski 2012 {published data only}
cognition, and walking in multiple sclerosis. Journal of
Dlugonski D, Motl RW, Mohr DC, Sandroff BM. Internet-delivered Neurology 2014;261(2):363-72.
behavioral intervention to increase physical activity in persons
with multiple sclerosis: sustainability and secondary outcomes.
Psychology Health and Medicine 2012;17(6):636-51. References to studies excluded from this review
Egner 2003 {published data only} Amato 2014 {published data only}
Egner A, Phillips VL, Vora R, Wiggers E. Depression, fatigue, Amato M, Goretti B, Viterbo R, Portaccio E, Niccolai C, Hakiki B,
and health-related quality of life among people with advanced et al. Computer-assisted rehabilitation of attention in patients
multiple sclerosis: results from an exploratory telerehabilitation with multiple sclerosis: results of a randomized, double-blind
study. NeuroRehabilitation 2003;18(2):125-33. trial. Multiple Sclerosis 2014;20(1):91-8.
Finlayson 2011 {published data only} Beckner 2010 {published data only}
Finlayson M, Preissner K, Cho C, Plow M. Randomized Beckner V, Howard I, Vella L, Mohr DC. Telephone-administered
trial of a teleconference-delivered fatigue management psychotherapy for depression in MS patients: moderating
program for people with multiple sclerosis. Multiple Sclerosis role of social support. Journal of Behavioral Medicine
2011;17(9):1130-40. 2010;33(1):47-59.
Frevel 2014 {published data only} Bombardier 2008 {published data only}
Frevel D, Mäurer M. Internet-based home training is capable to Bombardier CH, Cunniffe M, Wadhwani R, Gibbons LE, Blake KD,
improve balance in multiple sclerosis: a comparative trial with Kraft GH. The efficacy of telephone counseling for health
hippotherapy. European Journal of Physical and Rehabilitation promotion in people with multiple sclerosis: a randomized
Medicine 2015;51(1):23-30. [PUBMED: 24755773] controlled trial. Archives of Physical Medicine and Rehabilitation
2008;89(10):1849-56.
Gutíerrez 2013a {published data only}
Bombardier 2013 {published data only}
Gutíerrez RO, Galán Del Río F, Cano de la Cuerda R, Alguacil
Diego IM, González RA, Page JC. A telerehabilitation program Bombardier CH, Ehde DM, Gibbons LE, Wadhwani R,
by virtual reality-video games improves balance and postural Sullivan MD, Rosenberg DE, et al. Telephone-based physical
control in multiple sclerosis patients. NeuroRehabilitation activity counseling for major depression in people with
2013;33(4):545-54. multiple sclerosis. Journal of Consulting and Clinical Psychology.
2013;81(1):89-99.
Huijgen 2008 {published data only}
Cerasa 2013 {published data only}
Huijgen BC, Vollenbroek-Hutten MM, Zampolini M, Opisso E,
Bernabeu M, Van Nieuwenhoven J, et al. Feasibility of a home- Cerasa A, Gioia M C, Valentino P, Nistico R, Chiriaco C,
based telerehabilitation system compared to usual care: arm/ Pirritano D, et al. Computer-assisted cognitive rehabilitation
hand function in patients with stroke, traumatic brain injury of attention deficits for multiple sclerosis: a randomized trial
and multiple sclerosis. Journal of Telemedicine and Telecare. with fMRI correlates. Neurorehabilitation and Neural Repair
2008;14(5):249-56. 2013;27(4):284-95.
Motl 2011 {published data only} Fischer 2013 {published data only}
Motl RW, Dlugonski D, Wojcicki TR, McAuley E, Mohr DC. Internet Fischer A, Schroder J, Pottgen J, Lau S, Heesen C, Moritz S, et
intervention for increasing physical activity in persons with al. Effectiveness of an internet-based treatment programme for
multiple sclerosis. Multiple Sclerosis 2011;17(1):116-28. depression in multiple sclerosis: a randomized controlled trial.
Multiple Sclerosis 2013;19(11):350-1.
Paul 2014 {published data only}
Ghahari 2010 {published data only}
Paul L, Coulter EH, Miller L, McFadyen A, Dorfman J,
George GMP. Web-based physiotherapy for people Ghahari S, Leigh Packer T, Passmore AE. Effectiveness of an
moderately affected with multiple sclerosis; quantitative online fatigue self-management programme for people with
and qualitative data from a randomized, controlled pilot chronic neurological conditions: a randomized controlled trial.
study. Clinical Rehabilitation 2014;28(9):924-35. [DOI: Clinical Rehabilitation 2010;24(8):727-44.
10.1177/0269215514527995]
Miller 2011 {published data only}
Pilutti 2014 {published data only} Miller DM, Moore SM, Fox RJ, Atreja A, Fu AZ, Lee JC, et al. Web-
* Pilutti L, Dlugonski D, Sandroff B, Klaren R, Motl R. based self-management for patients with multiple sclerosis: a
Randomized controlled trial of a behavioral intervention practical, randomized trial. Telemedicine Journal and e-Health
targeting symptoms and physical activity in multiple sclerosis. 2011;17(1):5-13.
Multiple Sclerosis 2014;20(5):594–601.

Informed decisions.

Mohr 2000 {published data only} Bendixen 2008

Mohr DC, Likosky W, Bertagnolli A, Goodkin DE, Van Bendixen RM, Levy C, Lutz BJ, Horn KR, Chronister K, Mann WC.
der Wende J, Dwyer P, et al. Telephone-administered cognitive- A telerehabilitation model for victims of polytrauma.
behavioral therapy for the treatment of depressive symptoms in Rehabilitation Nursing 2008;33(5):215-20.
multiple sclerosis. Journal of Consulting and Clinical Psychology
2000;68(2):356-61. Bendixen 2009
Bendixen RM, Levy CE, Olive ES, Kobb RF, Mann WC. Cost
Mohr 2005 {published data only} effectiveness of a telerehabilitation program to support
Mohr DC, Hart SL, Julian L, Catledge C, Honos-Webb L, Vella L, chronically ill and disabled elders in their homes. Telemedicine
et al. Telephone-administered psychotherapy for depression. Journal and e-Health 2009;15(1):31-8.
Archives of General Psychiatry 2005;62(9):1007-14.
Brennan 2009
Mohr 2007 {published data only} Brennan DM, Mawson S, Brownsell S. Telerehabilitation:
Mohr DC, Hart S, Vella L. Reduction in disability in a randomized enabling the remote delivery of healthcare, rehabilitation, and
controlled trial of telephone-administered cognitive-behavioral self management. Studies in Health Technology and Informatics
therapy. Health Psychology 2007;26(5):554-63. 2009;145:231-48.
Moss-Morris 2012 {published data only} Chumbler 2012

Moss-Morris R, McCrone P, Yardley L, Van Kessel K, Wills G, Chumbler NR, Quigley P, Li X, Morey M, Rose D, Sanford J, et al.
Dennison L. A pilot randomised controlled trial of an Internet- Effects of telerehabilitation on physical function and disability
based cognitive behavioural therapy self-management for stroke patients: a randomized, controlled trial. Stroke
programme (MS Invigor8) for multiple sclerosis fatigue. 2012;43(8):2168-74.
Behaviour Research and Therapy 2012;50(6):415-21.
Constantinescu 2010
Solari 2004 {published data only} Constantinescu G, Theodoros D, Russell T, Ward E, Wilson S,
Solari A, Motta A, Mendozzi L, Pucci E, Forni M, Mancardi G, et al. Wootton R. Assessing disordered speech and voice in
Computer-aided retraining of memory and attention in people Parkinson's disease: a telerehabilitation application.
with multiple sclerosis: a randomized, double-blind controlled International Journal of Language & Communication Disorders
trial. Journal of the Neurological Sciences 2004;222(1):99-104. 2010;45(6):630-44.
Stuifbergen 2012 {published data only} Currell 2000

Stuifbergen AK, Becker H, Perez F, Morison J, Kullberg V, Todd A. Currell R, Urquhart C, Wainwright P, Lewis R. Telemedicine
A randomized controlled trial of a cognitive rehabilitation versus face to face patient care: effects on professional practice
intervention for persons with multiple sclerosis. Clinical and health care outcomes. Cochrane Database of Systematic
Rehabilitation 2012;26(10):882-93. Reviews 2000, Issue 2. [DOI: 10.1002/14651858.CD002098]
Wiles 2003 {published data only} Egger 1998

Wiles CM, Newcombe RG, Fuller KJ, Jones A, Price M. Use Egger M, Smith GD. Bias in location and selection of studies.
of videotape to assess mobility in a controlled randomized BMJ 1998;316(7124):61-6.
crossover trial of physiotherapy in chronic multiple sclerosis.
Clinical Rehabilitation 2003;17(3):256-63. Finkelstein 2008
Finkelstein J, Lapshin O, Castro H, Cha E, Provance PG. Home-
Zissman 2012 {published data only} based physical telerehabilitation in patients with multiple
Zissman K, Lejbkowicz I, Miller A. Telemedicine for multiple sclerosis: a pilot study. Journal of Rehabilitation Research and
sclerosis patients: assessment using Health Value Compass. Development 2008;45(9):1361-73.
Multiple Sclerosis 2012;18(4):472-80.
Forducey 2003
Forducey PG, Ruwe WD, Dawson SJ, Scheideman-Miller C,
Additional references McDonald NB, Hantla MR. Using telerehabilitation to promote
Amatya 2013 TBI recovery and transfer of knowledge. NeuroRehabilitation
2003;18(2):103-11.
Amatya B, Khan F, La Mantia L, Demetrios M, Wade DT. Non
pharmacological interventions for spasticity in multiple Giansanti 2008
sclerosis. Cochrane Database of Systematic Reviews 2013, Issue
Giansanti D, Macellari V, Maccioni G. Telemonitoring and
2. [DOI: 10.1002/14651858.CD009974.pub2]
telerehabilitation of patients with Parkinson's disease: health
Beer 2012 technology assessment of a novel wearable step counter.
Telemedicine Journal and e-Health 2008;14(1):76-83.
Beer S, Khan F, Kesselring J. Rehabilitation interventions
in multiple sclerosis: an overview. Journal of Neurology GRADEpro 2008 [Computer program]
2012;259(9):1994-2008.
Brozek J, Oxman A, Schunemann H. GRADEpro. Version 3.2 for
windows. GRADE working group, 2008.
Informed decisions.

Gutierrez 2013b Khan 2010b

Ortiz-Gutierrez R, Cano-de-la-Cuerda R, Galan-del-Rio F, Khan F, Pallant JF, Zhang N, Turner-Stokes L. Clinical practice
Alguacil-Diego IM, Palacios-Cena D, Miangolarra-Page JC. improvement approach in multiple sclerosis rehabilitation:
A telerehabilitation program improves postural control in a pilot study. International Journal of Rehabilitation Research
multiple sclerosis patients: a Spanish preliminary study. 2010;33(3):238-47.
International Journal of Environmental Research and Public
Health 2013;10(11):5697-710. Kurtzke 1982
Kurtzke JF, Gundmunsson KR, Bergmann S. Multiple sclerosis
Gøetzsche 1987 in Iceland: evidence of post war epidemic. Neurology
Gøetzsche PC. Reference bias in reports of drug trials. British 1982;32(2):143-50.
Medical Journal (Clinical Research ed.) 1987;295(6599):654-6.
Lai 2004
Hailey 2010 Lai JC, Woo J, Hui E, Chan WM. Telerehabilitation - a new
Hailey D, Roine R, Ohinmaa A, Dennett L. Evidence on the model for community-based stroke rehabilitation. Journal of
effectiveness of telerehabilitation applications. Institute of Telemedicine and Telecare 2004;10(4):199-205.
Health Economics and Finnish Office for Health Technology
Assessment 2010. [ISBN: 978-1-897443-87-3] Legg 2004
Legg L, Langhorne P, Outpatient Service Trialists. Rehabilitation
Hailey 2011 therapy services for stroke patients living at home: systematic
Hailey D, Roine R, Ohinmaa A, Dennett L. Evidence of benefit review of randomised trials. Lancet 2004;363(9406):352-6.
from telerehabilitation in routine care: a systematic review.
Journal of Telemedicine and Telecare 2011;17(6):281-7. Mc Donald 2001
McDonald WI, Compston A, Edan G, Goodkin D, Hartung HP,
Hammond 2000 Lublin FD, et al. Recommended diagnostic criteria for
Hammond SR, McLeod JG, Macaskill P, English DR. Multiple multiple sclerosis: guidelines from the International Panel
sclerosis in Australia: prognostic factors. Journal of Clinical on the diagnosis of multiple sclerosis. Annals of Neurology
Neuroscience 2000;7(1):16-9. 2001;50(1):121-7.
Higgins 2011 McCue 2010

Higgins JPT, Green S (editors). Cochrane Handbook for McCue M, Fairman A, Pramuka M. Enhancing quality of life
Systematic Reviews of Interventions Version 5.1.0 [updated through telerehabilitation. Physical Medicine and Rehabilitation
March 2011]. The Cochrane Collaboration, 2011. Available from Clinics of North America 2010;21(1):195-205.
www.cochrane-handbook.org.
McLean 2010
Houlihan 2011 McLean S, Chandler D, Nurmatov U, Liu JLY, Pagliari C,
Houlihan BV, Jette A, Paasche-Orlow M, Wierbicky J, Car J, et al. Telehealthcare for asthma. Cochrane
Ducharme S, Zazula J, et al. A telerehabilitation intervention Database of Systematic Reviews 2010, Issue 10. [DOI:
for persons with spinal cord dysfunction. American Journal of 10.1002/14651858.CD007717.pub2]
Physical Medicine & Rehabilitation 2011;90(9):756-64.
McLean 2011
Johansson 2011 McLean S, Nurmatov U, Liu JLY, Pagliari C, Car J, Sheikh A.
Johansson T, Wild C. Telerehabilitation in stroke care, a Telehealthcare for chronic obstructive pulmonary disease.
systematic review. Journal of Telemedicine and Telecare Cochrane Database of Systematic Reviews 2011, Issue 7. [DOI:
2011;17(1):1-6. 10.1002/14651858.CD007718.pub2]
Kairy 2009 MS Australia 2012

Kairy D, Lehoux P, Vincent C, Visintin M. A systematic review MS Australia. Practice for health professionals: Spasticity and
of clinical outcomes, clinical process, healthcare utilization multiple sclerosis (MS). www.msaustralia.org.au/sites/default/
and costs associated with telerehabilitation. Disability and files/spasticity.pdf. [ISBN: 978-0-9806637-4-7]
Rehabilitation 2009;31(6):427-47.
NMSS 2007
Khan 2007 National Multiple Sclerosis Society. Technology and MS:
Khan F, Turner-Stokes L, Ng L, Kilpatrick T, Amatya B. new survey finds technology plays a critical role in lives of
Multidisciplinary rehabilitation for adults with multiple people with multiple sclerosis, yet many are not using It to
sclerosis. Cochrane Database of Systematic Reviews 2007, Issue overcome disease-related challenges. www.prnewswire.com/
2. [DOI: 10.1002/14651858.CD006036.pub2] news-releases/video-new-survey-finds-technology-plays-
a-critical-role-in-lives-of-people-with-multiple-sclerosis-
Khan 2010a yet-many-are-not-using-it-to-overcome-disease-related-
Khan F, Gray O. Disability management and rehabilitation for challenges-58929337.html (accessed 28th March 2015).
persons with multiple sclerosis. Neural Regeneration Research
2010;5(4):301-9.
Informed decisions.

Pfleger 2010 Tousignant 2011

Pfleger CC, Flachs EM, Koch-Henriksen N. Social consequences Tousignant M, Moffet H, Boissy P, Corriveau H, Cabana F,
of multiple sclerosis (1): early pension and temporary Marquis F. A randomized controlled trial of home
unemployment, a historical prospective cohort study. Multiple telerehabilitation for post-knee arthroplasty. Journal of
Sclerosis 2010;16(1):121-6. Telemedicine and Telecare 2011;17(4):195-8.
Polman 2005 Trisolini 2010

Polman CH, Reingold SC, Edan G, Filippi M, Hartung HP, Trisolini M, Honeycutt A, Wiener J, Lesesne S. Global economic
Kappos L, et al. Diagnostic criteria for multiple sclerosis: 2005 impact of multiple sclerosis: a literature review. www.msif.org/
revisions to the "McDonald Criteria". Annals of Neurology wp-content/uploads/2014/09/ExecSummary_English.pdf
2005;58(6):840-6. (Accessed 28th March 2015).
Poser 1983 Van Tulder 2003

Poser CM, Paty DW, Scheinberg L, McDonald WI, Davis FA, Van Tulder MW, Furlan A, Bombardier C, Bouter L. Updated
Ebers GC, et al. New diagnostic criteria for multiple sclerosis: method guidelines for systematic reviews in the Cochrane
guidelines for research protocols. Annals of Neurology Collaboration Back Review Group. Spine 2003;28(12):1290-9.
1983;13(3):227-31.
Weinshenker 1989
Review Manager 2014 [Computer program] Weinshenker BG, Bass B, Rice GP, Noseworthy J, Carriere W,
The Nordic Cochrane Centre, The Cochrane Collaboration. Baskerville J, et al. The natural history of multiple sclerosis:
Review Manager (RevMan). Version 5.3. Copenhagen: The a geographically based study I. Clinical course and disability.
Nordic Cochrane Centre, The Cochrane Collaboration, 2014. Brain 1989;112(Pt 1):133-46.
Rogante 2010 WHO 2001

Rogante M, Grigioni M, Cordella D, Giacomozzi C. Ten years of World Health Organization (WHO). International Classification
telerehabilitation: a literature overview of technologies and of Functioning, Disability and Health (ICF). www.who.int/
clinical applications. NeuroRehabilitation 2010;27(4):287-304. classifications/icf/en/ (Accessed May 2014).
Russell 2011 WHO 2008

Russell TG, Buttrum P, Wootton R, Jull GA. Internet-based World Health Organization (WHO). Atlas: Multiple Sclerosis
outpatient telerehabilitation for patients following total knee Resources in the World 2008. WHO and Multiple Sclerosis
arthroplasty: a randomized controlled trial. Journal of Bone and International Federation, 2008. [ISBN 978 92 4 156375 8]
Joint Surgery. American Volume 2011;93(2):113-20.
Winters 2002
Sandroff 2014 Winters JM. Telerehabilitation research: emerging
Sandroff BM, Klaren RE, Pilutti LA, Dlugonski D, Benedict RH, opportunities. Annual Review of Biomedical Engineering
Motl RW. Randomized controlled trial of physical activity, 2002;4:287-320.
cognition, and walking in multiple sclerosis. Journal of
Neurology 2014;261(2):363-72.
References to other published versions of this review
Steel 2011
Khan 2013
Steel K, Cox D, Garry H. Therapeutic videoconferencing
Khan F, Amatya B, Kesselring J. Telerehabilitation for persons
interventions for the treatment of long-term conditions. Journal
with multiple sclerosis. Cochrane Database of Systematic
of Telemedicine and Telecare 2011;17(3):109-17.
Reviews 2013, Issue 5. [DOI: 10.1002/14651858.CD010508]
Theodoros 2008

Theodoros D, Russell T. Telerehabilitation: current perspectives. * Indicates the major publication for the study
Studies in Health Technology and Informatics 2008;131:191-209.

CHARACTERISTICS OF STUDIES
Characteristics of included studies [ordered by study ID]

Dlugonski 2012
Methods RCT, parallel group with wait-list controls; USA
Study period: one month period of July 2010

Funding source: not mentioned
Declaration of interest: not mentioned

Informed decisions.

Dlugonski 2012 (Continued)
Participants N = 45: treatment group = 22 and control = 23
Inclusion: Diagnosis of relapsing-remitting MS (RRMS); relapse-free in the past 30 days; Internet access;
willingness to complete questionnaires; wear pedometer during intervention period; being non-active,
defined as engaging in regular activity (30 minutes accumulated/day) on ≤ 2 days of the week during
previous 6 months; ability to ambulate with or without assistance (i.e., walking with or without a cane/
walker, but not a wheelchair or scooter); free of contraindication for physical therapy (e.g., no underly-
ing cardiovascular disease); physician approval for beginning a physical activity programme
Exclusion: not specified
Demographic characteristics:
Mean age 46.6 years (SD: 9.7 years), 86.7% women, mean time since diagnosis 9.4 years (SD: 7.8 years),
64.4% had at least college degree, 95.6% white, 62.2% employed and 73.3% married
Interventions (similar to Motl 2011)
Treatment group: Internet-delivered and social cognitive theory (self efficacy, outcome expectations,
impediments, and goal setting) based behavioural intervention supplemented with video coaching
for 12 weeks, which included text-based content supplemented by video and portable document for-
mat (PDF) files (i.e. multimedia). The intervention consisted of 4 essential modules: Getting Started
(benefits of physical activity and information for becoming more physically active), Planning for Suc-
cess (goal setting and feedback, outcome expectations, and self efficacy), Beating the Odds (barriers
and strategies of overcoming barriers, and social support), and Sticking with It (maintaining an active
lifestyle and physical activity relapse prevention), with 10 total Chapters. This was further supported by
automated e-mail announcements about new information, updates, and changes on the web-site
Additionally, 7 one-on-one web-based video coaching interactive sessions (5 - 10 minutes) using web-
cam were conducted (4 in the first month, 2 in second month and 1 in third month), by an experienced
doctoral student. The coaching sessions included discussions about progress towards goal achieve-
ments, content of website and adverse events
For goal-setting and self-monitoring purposes a pedometer, log book to record steps and computer
programme “Goal tracker” to upload weekly steps counts onto the website were provided
Control group: wait-list participants, who received the intervention materials after study completion
Outcomes Primary outcome: Physical activity: GLTEQ
Secondary outcome: Walking mobility: MSWS-12; QoL: MSIS-29; disease severity: PDDS; participant
satisfaction (Process evaluation questionnaire)
Assessment time points: Baseline, post-intervention (12 weeks) and 3 months
Notes This study follows an earlier study (see below Motl 2011) and evaluated the same cohort of participants
from a single database for similar intervention
Risk of bias
Bias Authors' judgement Support for judgement
Random sequence genera- Unclear risk Participants were paired based on their baseline level of activity (GLTEQ) and
tion (selection bias) neurologic disability (PDDS) score by the authors, then randomised using a
random number table
Allocation concealment High risk Not reported, as randomisation was performed pairwise, allocation conceal-
(selection bias) ment was unlikely

Informed decisions.

Dlugonski 2012 (Continued)
Blinding of participants High risk No blinding of participants and treating personnel
and personnel (perfor-
mance bias)
All outcomes
Blinding of outcome as- High risk No blinding of outcome assessors

sessment (detection bias)
All outcomes
Incomplete outcome data Low risk Overall, only 1 participant from control group dropped out. ITT analysis per-
(attrition bias) formed
All outcomes
Selective reporting (re- Low risk All prespecified (primary and secondary) outcomes reported
porting bias)
Other bias Unclear risk Selective participants: recruitment occurred through a database of self-volun-
teering persons for research

Egner 2003
Methods RCT, 3 parallel groups; USA
Study period: not mentioned

Funding source: grant from the Centers for Disease Control and Prevention, National Center on Birth
Defects and Developmental Disabilities, USA
Participants N = 27: Group 1 (video) = 9; Group 2 (telephone) = 11 and Group 3 (standard care) = 7
Inclusion: diagnosis of MS; experience of a recent functional setback in the disease process, such as a
severe exacerbating episode or an increase or start of chemotherapy treatment; EDSS score ≥ 7
Mean age 46.0 years (SD: 9.0 years), 63% women, 44% married, 37% African –Americans and mean EDSS
score of 7.8 (SD 0.6)
Interventions Treatment group (Groups 1 and 2): structured in-home education and counselling session delivered via
telephone or video by a rehabilitation nurse, which included individual rehabilitation education ses-
sions (structured review of skin care, nutrition, bowel and bladder routines, psychosocial issues and
any equipment needs, and referrals to mental health counsellors, physical therapists, or other health
professionals as needed. The same protocol was followed for the video and telephone groups with
video group trained in the use of the Plain Old Telephone System (POTS) units in their home which pro-
vided image and sound
Sessions: 30 - 40 minutes, weekly for a period of 5 weeks, then once every 2 weeks for 1 month.
Control group: usual care with regular follow-up offered by the rehabilitation facility
Outcomes Primary outcome: Fatigue: FSS; HRQOL: QWB; Depression: CES-D
Secondary outcome: none
Assessment time points: Baseline, 5 weeks during intervention, post-intervention (9 weeks) and every
month for 24 months

Informed decisions.

Egner 2003 (Continued)
Notes This study was part of a larger study of the impact of a telerehabilitation intervention on people with
severe mobility impairment, with people with spinal cord injuries and the prevention of pressure sores
as the primary group of interest of the project
Risk of bias
Random sequence genera- Unclear risk Participants were randomly assigned to 1 of 3 intervention groups: video, tele-
tion (selection bias) phone, or standard care. Further details not provided
Allocation concealment Unclear risk Not reported

(selection bias)

mance bias)
All outcomes

All outcomes
Incomplete outcome data Low risk No drop-outs in either group

(attrition bias)
All outcomes
porting bias)
Other bias Unclear risk Participant recruitment process and methodology not described in detail
No power calculation for the study
Small sample size with unbalanced allocation of participants to groups
ITT analysis not performed

Finlayson 2011
Methods RCT, 2-group time series design with a wait-list control group; USA
Study period: November 2007 to April 2009

Funding source: Field-Initiated Research Grant, National Institute of Disability and Rehabilitation Re-
search, USA
Declaration of interest: authors declared no conflict of interest
Participants N = 190: treatment group = 94 and control group = 96
Inclusion: living within the state of Illinois; diagnosis of MS; ≥ 18 years; functional English literacy; Fa-
tigue Severity Scale (FSS) score ≥ 4 (i.e. moderate to severe fatigue); weighted score of at least 12 on the
short version of the Blessed Orientation Memory Concentration test.
Exclusion: not provided

Informed decisions.

Finlayson 2011 (Continued)
Mean age 56 yrs (SD 9), 79% women, mean disease duration 15 yrs (SD 9 yrs), 88% white, 52% RRMS;
37% employed; 98% with education > 12 years
Interventions Treatment group: a 6-week group-based, teleconference-delivered (70-minute) fatigue management

programme, facilitated by a licensed OT
Control group: wait-list control group receiving treatment after 8 - 12 weeks
Outcomes Primary outcome: fatigue impact: FIS, fatigue severity: FSS; HRQoL:SF-36
Secondary outcome: self efficacy: ECQ
Assessment time points: Baseline, post-intervention (6 weeks) 3 months and 6 months
Notes No adverse events were identified during the trial
Risk of bias
Random sequence genera- Low risk Participants' randomisation completed by the statistician using a random per-
tion (selection bias) mutated block design with each block consisting 4 people
Allocation concealment Low risk Opaque envelopes were used and prepared in advance of recruitment. The en-
(selection bias) velopes were numbered sequentially and a statement indicating the allocation
(immediate or wait-list) was placed in each envelope
Blinding of participants Unclear risk Participants blinded to group allocation only and treating personnel not blind-
and personnel (perfor- ed
mance bias)
All outcomes

All outcomes
Incomplete outcome data High risk Overall 39 participants (20.5%) drop-out (17 in intervention, 22 in control
(attrition bias) group)
All outcomes
porting bias)
Other bias Unclear risk ITT analysis performed for effectiveness analysis

Frevel 2014
Methods RCT, parallel group; Germany

Inclusion: Definite MS diagnosis according to McDonald's criteria, EDSS 2-6, ability to stand with or
without an assistive device for 1 minute, age 18 - 60 years, clinical stability for last 4 weeks

Informed decisions.

Frevel 2014 (Continued)
Exclusion: clinically relevant internal or orthopaedic diseases unrelated to MS, an allergy or aversions
to horses or previous experience with hippotherapy or therapeutic ridings (since diagnosis of MS)
Mean age 45.5 years (range 32 - 57), mean EDSS 3.8 (range 2 – 6), mean disease duration 19.0 (range 1 -
35), RRMS 67%
Interventions Treatment group; Internet-based home training: balance, postural control exercises and strength train-
ing for main group of muscles of the lower extremities, trunk and shoulder griddle. Participant provid-
ed feedback (Borg scale) to the therapist, which provided further feedback after each sessions (dura-
tion 2 training sessions (45 minutes)/week for 12 weeks). Further, participants had an informative su-
pervised meeting and received instructions and software prior
Control group: hippotherapy twice per week/ 20 – 30 minutes under supervision of riding therapist for
12 weeks
Outcomes Primary Outcomes: Balance: BBS, DGI
Secondary outcomes: Isometric muscle strength of knee and trunk; TUG; 2MWT; HAQUAMS, FSS, MFIS
Assessment time points: Baseline and post intervention (12 weeks)
Notes No report of adverse events
Risk of bias
Random sequence genera- Low risk Randomised by simple allocation by drawing lots of preshuffled opaque en-
tion (selection bias) velopes
Allocation concealment Low risk Sealed opaque envelopes containing an identifier were used
(selection bias)

mance bias)
All outcomes

All outcomes
Incomplete outcome data Low risk Overall, 2 participants in treatment group dropped out (11%)
(attrition bias)
All outcomes
porting bias)
Other bias Unclear risk No power calculations for the study
No ITT analysis
Small sample size

Informed decisions.


Gutíerrez 2013a
Methods RCT, parallel group; Spain

Participants Spain. N = 50: treatment group = 25 and control group = 25
Inclusion: Confirmed diagnosis of MS for > 2 years based on McDonald's criteria; age 20 - 60 years; med-
ically stable during 6 months prior to baseline assessment; impaired balance demonstrated by MRI;
EDSS score of 3 - 5; Hauser ambulatory index > 4, absence of cognitive impairment (MMSE ≥ 24); no vi-
sual deficit; internet connection at home.
Exclusion: diagnosis with other disease or pathological condition that affects balance; had a relapse in
the month prior to baseline or during the intervention process; received intravenous or oral steroid cy-
cle prior to beginning the evaluation protocol and within 4-month duration of intervention
Treatment group: Mean age 39.7 years (SD 8.1), 54% women, mean disease duration 9.7 years (SD 6.8),
EDSS score ≥ 4: 83.6%, RR MS: 71.9%
Control group: mean age 42.8 years (SD 7.4), 61% women, mean disease duration 10.9 years (SD 5.4),
EDSS score ≥ 4: 78.3%, RR MS: 65.2%
Interventions Treatment group: monitored virtual reality telerehabilitation programme via video-conference us-
ing the Xbox 360® and Kinect console, which included gaming protocol consisted of 3 games (Kinect
Sports, Kinect Joy Ride, and Kinect Adventures).proposing activities that involve integrating proprio-
ceptive, visual, and vestibular sensory information. Responses directed to the maintenance of balance
and postural stability are triggered by the visual feedback that participants continuously receive in re-
al time with regard to their position, performance type, and the movement direction that the task re-
quires. The protocol proposed tasks such as throwing and hitting objects with one’s hands and feet,
hitting and receiving balls with different body parts, dodging objects, overcoming obstacles, imitating
postures, or managing virtual elements that favour key aspects of postural control (e.g., girdle dissoci-
ation, alternating load distribution, changes in direction, multidirectional movement, reaction speed,
hand-eye co-ordination, foot-eye co-ordination, and dexterity) in different positions across a stepwise
gradient of difficulty. Experimental group attended 40 sessions, 4 sessions per week (20 minutes per
session) at home
Control group: Ambulatory PT twice/week for 10 weeks (40 minutes per session) at rehab centre
Outcomes Primary outcome: Postural control : CDP; SOT; motor function: MCT
Secondary outcome: clinical outcomes: BBS, TS
Assessment points: Baseline and post-intervention (10 weeks)
Short-term follow-up
Same study published in different journals by the same authors (Gutierrez 2013b)
Risk of bias
Random sequence genera- High risk Participants allocated to treatment or control groups based on the specific
tion (selection bias) criteria. Only after screening for the treatment group, remaining participants
were randomly distributed into 2 groups using computer software. Further,

Informed decisions.

Gutíerrez 2013a (Continued)
2 participants were added to the treatment group due to availability of the
equipment
Allocation concealment High risk No allocation concealment

(selection bias)

mance bias)
All outcomes
Blinding of outcome as- Low risk Outcome assessors were blinded

All outcomes
Incomplete outcome data Low risk 3 drop-outs ( 1 in treatment group and 2 in control group)
(attrition bias)
All outcomes
porting bias)
Other bias High risk No power calculation
No ITT analysis
Small sample size

Huijgen 2008
Methods RCT, parallel group, multicentred; Italy, Spain and Belgium
Study period: October 2005 to January 2007

Funding source: study was part of a project supported by European Union
Participants N = 81 (Stroke = 16, TBI = 30, MS = 35): treatment group = 55 (MS = 24) and control = 26 (MS = 11)
Inclusion: age > 18 years; confirmed diagnosis of MS, stroke or TBI; Nine Hole Peg Test > 25 sec and abil-
ity to move at least 1 peg in 180 sec; sufficient autonomous functioning; Internet connection or tele-
phone line and reachable Internet provider; stable clinical status and living at home
Exclusion: disturbed upper limb function not related to MS, stroke or TBI; serious cognitive and/or be-
havioural problems; serious emotional problems; major visual problems; communication problems;
medical complications; other problems possibly contraindicating autonomous exercise at home
Intervention group: mean age: 47 years (SD 18) (MS 48 years (SD 12)), 71% men (MS 46% men), mean
disease duration 9.7 years (SD 7.8 years) (MS 15.1 years (SD 8.6));
Control group: mean age: 50 years (SD 18) (MS 51 years (SD 14)), 69% men (MS 64% men), mean disease
duration 10.2 years (SD 7.6 years) (MS 15.6 years (SD 7.8))
Interventions Treatment group: 1 month of usual care followed by the Home Care Activity Desk (HCAD) – a telere-
habilitation intervention for arm/hand function at home which consisted a set of exercises for correct
functional activity of the upper limb such as reaching, grasping, lateral pinch, pinch grip, holding, ma-
nipulation and finger dexterity; and additional features for videoconferencing and recording. HCAD sys-

Informed decisions.

Huijgen 2008 (Continued)
tem comprised a hospital-based server and portable unit installed at participant’s home. At least 1 ses-
sion (30 minutes)/day for 5 days per week for 1 month
Control group: Usual care and generic exercises prescribed by their physicians
Outcomes Primary outcome: Upper limb function : ARAT; NHPT
Secondary outcome: participant satisfaction (VAS)
Heterogeneous in approach and intensity for control group activities
Higher percentage of men in the control group
Risk of bias
Random sequence genera- Unclear risk Participants randomly allocated to treatment or control group, in such way to
tion (selection bias) fit the clinical practice in a 2:1 ratio

(selection bias)

mance bias)
All outcomes

All outcomes
Incomplete outcome data Unclear risk Overall 11 participants (14%) were lost to follow-up (7 in intervention, 4 in con-
(attrition bias) trol group). Percentage of drop-outs reported but not time points
All outcomes
porting bias)
Other bias Unclear risk Study was underpowered

Motl 2011
Methods RCT, parallel group, with wait-list control; USA

Funding source: none
Inclusion: Definite diagnosis of RRMS; independently ambulatory or ambulatory with single-point assis-
tance (i.e. cane); relapse-free in the past 30 days; Internet access; willingness to complete the question-
naires and undergo randomisations; being non-active defined as not engaging in regular physical activ-
ity (30 minutes accumulated per day) on more than 2 days of the week during the previous 6 months;

Informed decisions.

Motl 2011 (Continued)
free of contraindications for physical activity (e.g. no underlying cardiovascular disease); and physician
approval for beginning a physical activity programme
Intervention group: mean age:46.1 years (SD 10.4), 90% women; mean disease duration: 8.1 years (SD
6.5); mean Determined Disease Steps Scale score (disease severity): 2.0 (SD 1.8)
Control group: mean age 45.6 (SD 9.2), 88% women, mean disease duration: 7.3 (SD 6.2), mean Deter-
mined Disease Steps Scale score (disease severity): 2.1 (1.9)
Interventions Treatment group: Internet intervention based on social cognitive theory (self efficacy, outcome expec-
tations, impediments, and goal setting), which included text-based content supplemented by video
and portable document format (PDF) files (i.e. multimedia). It consisted of 4 essential modules: Get-
ting Started (benefits of physical activity and information for becoming more physically active), Plan-
ning for Success (goal setting and feedback, outcome expectations, and self efficacy), Beating the Odds
(barriers and strategies of overcoming barriers, and social support), and Sticking with It (maintaining
an active lifestyle and physical activity relapse prevention), with 10 total Chapters. Additionally, inter-
active sessions twice per week were conducted, which included an ongoing participant forum for dis-
cussions of physical activity behaviour change, and a toll-free telephone line and a study e-mail ad-
dress for supporting the website. This was further supported by automated e-mail announcements
about new information, updates, and changes on the website
Control group: wait-list participants, who received the intervention materials after study completion
Outcomes Measured at baseline, immediately post-treatment (12 weeks after start of intervention)
Primary outcome: Physical activity: GLTEQ; Self efficacy: EXSE; Outcome expectations: MOEES; Func-
tional limitations: - Functional Limitations component of the abbreviated LL-FDI; Goal setting: EGS
Secondary outcome: Disease severity: PDDS
Assessment time points: Baseline and post-intervention (1 month)
Risk of bias
Random sequence genera- Unclear risk Participants were initially paired on physical activity and neurological disabili-
tion (selection bias) ty levels by 2 authors and then members of the pairs were randomly assigned
into intervention or wait-list control conditions

(selection bias)

mance bias)
All outcomes

All outcomes
Incomplete outcome data Unclear risk Overall 10 participants (15%) dropped out (6 in intervention, 4 in control
(attrition bias) group). Percentage of drop-outs reported but not time points
All outcomes

Informed decisions.

Motl 2011 (Continued)
porting bias)
Other bias Low risk None

Paul 2014
Methods RCT, parallel group; Scotland, UK

Funding source: grant, the Chief of Scientist Office, Scotland
Inclusion: Confirmed diagnosis of MS, EDSS: 5 - 6, stable drug therapy for 30 days, no relapses in the
previous 3 months, no significant comorbidities (such as co-existing cardiac or pulmonary condition),
have access to the Internet via personal or tablet computer
Further inclusion in the treatment group if participants did not receive conventional physiotherapy
treatment based on at least 1 the following criteria: (a) time on the waiting list; (b) limited geographic
accessibility; (c) unable to reconcile working hours and therapy schedule; or d) dependent on others to
arrive at the treatment centre
Treatment group: Mean age 50.8 years (SD 7.4), 80% women; mean disease duration 12.5 years (SD
7.1), mean EDSS 6. 0 (SD 0.5)
Control group: Mean age 52.5 years (SD 14.3), 80% women; mean disease duration 12.8 years (SD 10.9),
mean EDSS 5.8 (SD 0.5)
Interventions Treatment group: 12 weeks of individualised web-based physiotherapy completed twice per week. The
website consisted of a home page, exercise pages and advice section. Each exercise page contained a
video and text explaining the exercise, an audio description of the exercise and a timer. The catalogue
of exercises consisted of: cardiovascular, strengthening and balance exercises, each at 4 levels of diffi-
culty, as well as warm-up and cool-down exercises and stretches
Control group: usual care
Outcomes Primary outcomes: 25 Foot Walk Test
Secondary outcomes: BBS, TUG, MSIS, LMSQOLS, MS-Related Symptom Checklist, HADS, feasibility
and satisfaction with the programme
Assessment points: Baseline and post-intervention (12 weeks)
Short-term follow-up
Risk of bias
Random sequence genera- Low risk Randomisation performed using a series of random numbers, generated in
tion (selection bias) Microsoft Excel. Recruited participants were allocated consecutive numbers,

Informed decisions.

Paul 2014 (Continued)
where even numbers represented the intervention group and odd numbers
the control group

(selection bias)

mance bias)
All outcomes

All outcomes
Incomplete outcome data Low risk Overall, 1 participant dropped out from control group
(attrition bias)
All outcomes
porting bias)
Other bias Unclear risk Underpowered study
No ITT analysis
Small sample size

Pilutti 2014
Methods RCT, parallel group with wait-list controls, USA

Funding source: various grant from the National Multiple Sclerosis Society, USA; the Multiple Sclerosis
Society of Canada and the Multiple Sclerosis International Federation
Inclusion: 18 – 64 years; definite diagnosis of MS based on physician verification; relapse-free for the
past 30 days; Internet access; and ability to walk with or without an assistive device; physician’s ap-
proval for participation; willing and able to travel to the research site; have minimal risk for engaging in
physical activity (i.e. reported ‘yes’ to fewer than 2 questions on the PARQ)
Exclusion: participants who self-reported accumulating ≥ 30 minutes of moderate-to-vigorous physical

activity per day on ≥ 2 days/week.
Treatment group: Mean age 48.4 years (SD: 9.1 years), 73.2% women, mean time since diagnosis 10.6
years (SD: 7.1 years), RRMS 75.6%, PDSS: median 2.0 (IQR 4, 0)
Control group: Mean age 49.5 years (SD: 9.2 years), 78% women, mean time since diagnosis 13.0 years
(SD: 9.1 years), RRMS 83%, PDSS: median 3.0 (IQR 3, 0)
Interventions Treatment group: same as in Dlugonski 2012, Motl 2011 (see above). In addition, participant wore a
Yamax SW-401 Digiwalker pedometer, completed a log book and used Goal Tracker software, and re-
ceived a web-cam, and website information. Participants participated in 15 scheduled one-on-one
video coaching sessions for 6 months.

Informed decisions.

Pilutti 2014 (Continued)
Control group: wait-list participants, who received the intervention materials after study completion.
Outcomes Primary outcome: Physical activity: GLTEQ; fatigue: FSS, MFIS; depression and anxiety: HADS; pain:
MPQ; sleep: PSQI; HRQoL: MSIS-29, Cognitive processing speed: SDMT
Secondary outcome: disease severity:PDDS
Assessment time points: baseline and post-intervention (6 months)
Notes This RCT was considered the primary study, whose results were described in 2 different articles report-
ing different outcomes (Sandroff 2014) .
This study is part of a series of studies conducted earlier (Dlugonski 2012 and Motl 2011).
Risk of bias
Random sequence genera- Unclear risk After baseline testing, participants were grouped into matched pairs based on
tion (selection bias) step counts from the accelerometer and level of disability, and then random-
ly assigned to either the intervention or wait-list control condition using a ran-
dom numbers sequence
Allocation concealment High risk Not reported

(selection bias)

mance bias)
All outcomes

All outcomes
Incomplete outcome data Low risk Overall, 6 participants (7%) (4 from intervention and 2 from control group)
(attrition bias) dropped out
All outcomes
porting bias)
Other bias Unclear risk Selective participants: recruitment occurred through a database of self-volun-
teered persons for research
No ITT analysis performed (analysis of completers only)
USD 50 remuneration given to participants for completing each testing session
ARAT: Action Research Arm Test; BBS: Berg Balance Scale;CCT: Controlled clinical trial; CDP: Computerized Dynamic Posturography;
CES: Composite Equilibrium Score; CES-D: Center for Epidemiologic Studies Depression Scale; CI: Confidence interval;DGI: Dynamic
gait Index;ECQ: Energy Conservation Questionnaire; EDSS: Expanded Disability Status Scale; EGS: Exercise Goal setting Scale;ES: Effect
size; EXSE: Exercise Self-Efficacy Scale; FIS: Fatigue Impact Scale,), FSS: Fatigue Severity Score; GLTEQ: Godin Leisure-Time Exercise
Questionnaire; HADS: Hospital Anxiety and Depression Scale; HAQUAMS: Hamburg QoL Questionnaire in MS; HCAD: Home Care Activity
Desk; HRQoL: Health related quality of life; IPAQ: International Physical Activity Questionnaire; IQR: inter quartile range;ITT: intention to
treat; LMSQOLS: Leeds MS Quality of Life Scale;LL-FDI: Late-Life Function and Disability Instrument; MCT: Motor Control Test; MOEES:
Multidimensional Outcomes Expectations for Exercise Scale; MPQ: McGill Pain Questionnaire; MRI: Magnetic Resonance Imaging: MS:
Multiple Sclerosis;MSIS-29: MS Impact Scale; MSWS-12: MS Walking Scale – 12; NHPT: Nine Hole Peg Test; PARQ: Physical Activity
Readiness Questionnaire; PDDS: Patient Determined Disease Steps; PSQI: Pittsburgh Sleep Quality Index; QoL: quality of life; QWB: Quality

Informed decisions.

of Well- Being Scale; RCT: randomised controlled trial; RR: Risk Ratio; SD: Standard deviation; SDMT: Symbol Digit Modalities Test; SE:
Standard Error; SF-36: 36-Item Short Form Health Survey; SOT: Sensory organisation Test; TBI: traumatic brain injury; TS: Tineti Scale;
TUG: Timed Up and Go;UK: United Kingdom;USA: United States of America; VAS: Visual Analogue Scale; 2MWT: 2 minute walk test;6MWT:
6 minute walk test; 25FWT: 25 Foot Walk Test

Characteristics of excluded studies [ordered by study ID]

Study Reason for exclusion
Amato 2014 Not intervention of interest (mental health care only)
Beckner 2010 Not intervention of interest (mental health care only)
Bombardier 2008 Not Intervention of interest (telephone counselling for health promotion)
Bombardier 2013 Not Intervention of interest (telephone counselling for major depression)
Cerasa 2013 Not intervention of interest (mental health care only)
Fischer 2013 Not intervention of interest (mental health care only)
Ghahari 2010 No subgroup analysis for MS participants
Miller 2011 Intervention with no rehabilitation objectives
Mohr 2000 Not intervention of interest (mental health care only)
Moss-Morris 2012 Not intervention of interest (mental health care only)
Solari 2004 Not intervention of interest (mental health care only)
Stuifbergen 2012 Not intervention of interest (mental health care only)
Wiles 2003 Intervention: no telerehabilitation
Zissman 2012 Not intervention of interest (medical care only)

ADDITIONAL TABLES

Table 1. List of outcome measures used in the included studies*
Outcome Measures
Function
Action Research Arm Test (ARAT)
Berg Balance Scale (BBS)
Computerized Dynamic Posturography (CDP)

Informed decisions.

Table 1. List of outcome measures used in the included studies* (Continued)

Composite Equilibrium Score (CES)
Dynamic gait Index (DGI)
Exercise Self-Efficacy Scale (EXCE)
Godin Leisure-Time Exercise Questionnaire (GLTEQ)
International Physical Activity Questionnaire (IPAQ)
Late-Life Function and Disability Instrument (LL-FDI)
Motor Control Test (MCT)
Multidimensional Outcomes Expectations for Exercise Scale (MOEES)
Multiple Sclerosis Walking Scale – 12 (MSWS-12)
Nine-Hole Peg Test (NHPT)
Physical Activity Readiness Questionnaire (PARQ)
Sensory organisation Test (SOT)
Six Meter Walk Test (6MWT)
Tineti Scale (TS)
Timed Up and Go (TUG)
Two Meter Walk Test (2MWT)
Twenty-five Foot Walk Test (25-FWT)
Visual Preference Ratio (VPR)
Impairment and symptoms
Fatigue impact scale (FIS)
Fatigue severity scale (FSS)
Modified Fatigue impact scale (MFIS)
McGill Pain Questionnaire (MPQ)
MS related symptom check list
Pittsburgh Sleep Quality Index (PSQI)
Participation
Quality of Life
Hamburg Quality of Life Questionnaire in Multiple Sclerosis (HAQUAMAS)

Informed decisions.

Table 1. List of outcome measures used in the included studies* (Continued)

Leeds Multiple Sclerosis Quality of Life Scale (LMSQOL)
Multiple Sclerosis Impact Scale (MSIS-29)
Quality of Well- Being Scale (QWB)
36 item Short Form Health Survey Questionnaire (SF 36)
Psychological
Centre for Epidemiologic Studies Depression Scale (CES-D)
Hospital Anxiety and Depression Scale (HADS)
Symbol Digit Modalities Test (SDMT)
Other
Expanded Disability Status Scale (EDSS)
Energy Conservation Questionnaire (ECQ)
Exercise Goal setting Scale (EGS)
Muscle strength
Patient Determined Disease Steps (PDDS)
Self-Efficacy for Energy Conservation (SEEC)
Satisfaction with the intervention
Visual Analogue Scales (VAS)
*Outcome measures are categorised according to the International Classification of Functioning, Disability and Health (ICF, WHO 2001)

Table 2. Summary of telerehabilitation interventions in the included studies (Continued)
Study Telerehabilitation interventions
Contents Settings Technology Duration/intensity
Dlugonski Same as Motl 2011 ( see below) Partic- Internet-deliv- 12 weeks Same as Motl
2012 ipants' ered 2011 ( see below)
home
Egner Structured in-home education and counselling session Partic- Telephone or 30 to 40 minutes, week-
2003 delivered by a rehabilitation nurse, which included indi- ipants' video ly for 5 weeks, then
vidual rehabilitation education sessions home once every 2 weeks for 1
month.
Finlayson Group-based fatigue management programme, facilitat- Rehab Teleconference 70-minute weekly for 6
2011 ed by a licensed Occupational Therapist centre weeks

Informed decisions.

Table 2. Summary of telerehabilitation interventions in the included studies (Continued)

Frevel Training programme: balance, postural control exercises Partic- Internet-deliv- 2 training sessions/(45
2014 and strength training with additional interactive sessions ipants' ered minutes) weekly for 12
home weeks
Gutíerrez Monitored telerehabilitation programme, which includ- Partic- Virtual reali- 40 sessions, 4 sessions
2013a ed gaming protocol, proposing activities that involve inte- ipants' ty system via per week (20 minutes per
grating proprioceptive, visual, and vestibular sensory in- home video-confer- session)
formation. Experimental group attended at home ence using the
Xbox 360 and
Kinect console
Huijgen Home Care Activity Desk (HCAD) – a telerehabilitation in- Partic- Virtual telere- 1 month of usual care fol-
2008 tervention for arm/hand function and additional features ipants' habilitation lowed by HCAD- 1 ses-
for videoconferencing and recording. HCAD system home programme sion (30 minutes)/day
and video-con- for 5 days per week for 1
ference, com- month
prising a hospi-
tal-based serv-
er and portable
unit installed
at participant’s
home
Motl 2011 Same as Dlugonski 2012 (see above) Partic- Internet-deliv- Same as Dlugonski 2012
ipants' ered (see above)
home
Paul 2014 Individualised physiotherapy programme consisting of ex- Partic- Internet-deliv- Twice per week for 12
ercise page containing a video and text explaining the ex- ipants' ered weeks
ercise, an audio description of the exercise and a timer home
Pilutti Same as in Motl 2011 (see above), in addition, participant Partic- Internet-deliv- 15 scheduled one-on-one
2014 wore a Yamax SW-401 Digiwalker pedometer, completed ipants' ered video coaching sessions
a log book and used Goal Tracker software, and received a home for 6 months
web-cam, and website information
Sandroff Same as in Motl 2011, Pilutti 2014 (see above). In addi- Partic- Internet-deliv- Weekly one-on-one
2014 tion, website materials were delivered in a titrated man- ipants' ered behavioural coaching
ner over the 6-month period such that new content be- home sessions via Skype (15
came available 7 times during the first 2-month period, 4 scheduled sessions) for 6
times during the second 2-month period, and twice during months
the final 2 months of the intervention.

Table 3. Summary of outcome assessed in the included studies (Continued)
Study Outcome assessed*
Function Impairment Participation Others
Dlugonski 2012 GLTEQ, MSWS-12 MSIS-29 PDDS, SATISFACTION
Egner 2003 FSS QWB, CES-D
Finlayson 2011 FIS, FSS SF-36 ECQ, PDDS

Informed decisions.

Table 3. Summary of outcome assessed in the included studies (Continued)

Frevel 2014 BBS, DGI, TUG, 2MWT MFIS HAQUAMAS Muscle strength
Gutíerrez 2013a SOT, MCT, BBS, TS
Huijgen 2008 ARAT, NHPT VAS satisfaction sur-

vey
Motl 2011 GLTEQ, LL-FDI, EXCE, MOEES EGS, PDSS
Paul 2014 25 FWT, BBS, TUG MS related symp- MSIS, LMSQOL,

tom check list HADS
Pilutti 2014 GLTEQ MFIS, FSS, MPQ, MSIS-29, HADS PDDS

PSQI
Sandroff 2014 6MWT, IPAQ SDMT PDDS
*Categorised according to the International Classification of Functioning, Disability and Health (ICF, WHO 2001)
ARAT: Action Research Arm Test; BBS: Berg Balance Scale;CDP: Computerized Dynamic Posturography; CES: Composite Equilibrium
Score; CES-D: Center for Epidemiologic Studies Depression Scale; DGI: Dynamic gait Index;ECQ: Energy Conservation Questionnaire;
EDSS: Expanded Disability Status Scale; EGS: Exercise Goal setting Scale;EXSE: Exercise Self-Efficacy Scale; FIS: Fatigue Impact Scale;
FSS: Fatigue Severity Score; GLTEQ: Godin Leisure-Time Exercise Questionnaire; HADS: Hospital Anxiety and Depression Scale; HAQUAMS:
Hamburg Quality of Life Questionnaire in Multiple Sclerosis; IPAQ: International Physical Activity Questionnaire; LMSQOLS: Leeds Multiple
Sclerosis Quality of Life Scale;LL-FDI: Late-Life Function and Disability Instrument; MCT: Motor Control Test; MOEES: Multidimensional
Outcomes Expectations for Exercise Scale; MPQ: McGill Pain Questionnaire; MSIS-29: Multiple Sclerosis Impact Scale; MSWS-12: Multiple
Sclerosis Walking Scale–12; NHPT: Nine Hole Peg Test; PARQ: Physical Activity Readiness Questionnaire; PDDS: Patient Determined
Disease Steps; PSQI: Pittsburgh Sleep Quality Index; QWB: Quality of Well- Being Scale; SDMT: Symbol Digit Modalities Test; SF-36: 36-Item
Short Form Health Survey; SOT: Sensory organisation Test; TS: Tineti Scale; TUG: Timed Up and Go;VAS: Visual Analogue Scale; 6MWT:
6 minute walk test; 25FWT: 25 Foot Walk Test

APPENDICES
Appendix 1. Keywords
{telecommunications*} OR {telemedicine} OR {telehealth} OR {telehealthcare} OR {telecoahing} OR {e-health} OR {e-medicine} OR {mobile
health} OR {information technology} OR {information communication technology} OR {internet} OR {web-based} OR {computer} OR
{Software} OR {videoconferencing} OR {remote consultation} OR {remote sensing technology} OR {rehabilitation} OR {physiotherapy} OR
{occupational therapy} OR {speech therapy} OR {dietician}
CONTRIBUTIONS OF AUTHORS
Fary Khan (FA), and Bhasker Amatya (BA) were involved in all aspects of the review. Jurg Kesselring (JK) provided valuable input into design
of the review. Fary Khan, Bhasker Amatya, Mary Galea (MG) were responsible for all study selection, data extraction and methodological
quality of included studies. M Galea and J Kessering also provided valuable assistance with the Discussion. All review authors critically
reviewed the manuscript and discussed data collection, results and conclusions.
DECLARATIONS OF INTEREST
The review authors are clinicians and researchers in the field of Physical and Medical Rehabilitation who wish to provide the best possible
service to their patients.
Fary Khan: none known.
Bhasker Amatya: none known.
Jurg Kesselring: none known.
Mary Galea: none known.

Informed decisions.

SOURCES OF SUPPORT
Internal sources
• Department of Rehabilitation Medicine, Royal Melbourne Hospital, Australia.
External sources
• None, Other.
DIFFERENCES BETWEEN PROTOCOL AND REVIEW
We have included a 'Summary of findings' table in the review with the key outcomes identified categorised according to the WHO ICF
framework, which the authors deemed to be the most relevant to decision-makers including patients, clinicians and policy makers.
We have clarified ‘Types of interventions’ in this review to include control conditions: “any type of traditional face-to face rehabilitation
treatment in outpatient or day treatment settings”.
We exclude studies if they investigated interventions related to: “telerehabilitation targeting mental health conditions or substance abuse”;
“home care (or tele-home care) with no rehabilitation objectives”; “satisfaction with or acceptance of telerehabilitation technology” and
“technical development or feasibility of telerehabilitation”.
We modified ‘Data extraction and management’ for the review and added the following statement: “Data were extracted for intention-to-
treat (ITT) analysis from each study and where ITT data were not available, 'on-treatment’ data or the data of those who completed the
trial were retrieved.”
Based on the findings, we did not implement the planned methods as described in the protocol related to assessment of heterogeneity,
assessment of reporting bias, and data synthesis.
INDEX TERMS
Medical Subject Headings (MeSH)

*Telemedicine; Multiple Sclerosis [*rehabilitation]; Randomized Controlled Trials as Topic; Treatment Outcome
MeSH check words

Adult; Humans; Middle Aged


Khan Et Al-2015-Cochrane Database of Systematic Reviews PDF

Uploaded by

Copyright:

Available Formats

Khan Et Al-2015-Cochrane Database of Systematic Reviews PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Khan Et Al-2015-Cochrane Database of Systematic Reviews PDF

Uploaded by

Copyright:

Available Formats

Khan F, Amatya B, Kesselring J, Galea M

Khan F, Amatya B, Kesselring J, Galea M.

Telerehabilitation for persons with multiple sclerosis (Review) i

Telerehabilitation for persons with multiple sclerosis

Fary Khan1,2,3, Bhasker Amatya1, Jurg Kesselring4, Mary Galea5

Data collection and analysis

Telerehabilitation for persons with multiple sclerosis (Review) 1

Telerehabilitation for persons with multiple sclerosis

Telerehabilitation for persons with multiple sclerosis

Patient or population: People with multiple sclerosis

Settings: Participants' home, MS regional centres

Change in functional activity

Telerehabilitation for persons with multiple sclerosis (Review) 3

Change in symptoms or impairments

Telerehabilitation for persons with multiple sclerosis (Review) 4

Telerehabilitation for persons with multiple sclerosis (Review) 5

One study (Paul 2014, N = 30) reported non-significant improvement in HRQoL mea-

Change in other outcomes

# GRADE Working Group grades of evidence

Telerehabilitation for persons with multiple sclerosis (Review) 7

BACKGROUND method to deliver rehabilitation treatment in a setting convenient

offered an opportunity to provide an individualised rehabilitation Types of interventions

OBJECTIVES • acute medical/surgical/pharmacological interventions for

Telerehabilitation for persons with multiple sclerosis (Review) 9

Telerehabilitation for persons with multiple sclerosis (Review) 10

We considered a study to be of 'high' methodological quality if the Data synthesis

Telerehabilitation for persons with multiple sclerosis (Review) 12

Figure 1. Study flow diagram.

Telerehabilitation for persons with multiple sclerosis (Review) 13

Telerehabilitation for persons with multiple sclerosis (Review) 14

Telerehabilitation for persons with multiple sclerosis (Review) 15

Telerehabilitation for persons with multiple sclerosis (Review) 16

Telerehabilitation for persons with multiple sclerosis (Review) 18

Quality of life No studies reported data on cost effectiveness, investment costs

Egner 2003 reported no significant difference in QoL between the DISCUSSION

Telerehabilitation for persons with multiple sclerosis (Review) 19

Telerehabilitation for persons with multiple sclerosis (Review) 20

Implications for practice Telerehabilitation has a major role in providing remote

• More emphasis on participatory domains (cognitive outcomes ACKNOWLEDGEMENTS

Telerehabilitation for persons with multiple sclerosis (Review) 22

Telerehabilitation for persons with multiple sclerosis (Review) 23

Mohr 2000 {published data only} Bendixen 2008

Moss-Morris 2012 {published data only} Chumbler 2012

Stuifbergen 2012 {published data only} Currell 2000

Wiles 2003 {published data only} Egger 1998

Gutierrez 2013b Khan 2010b

Higgins 2011 McCue 2010

Kairy 2009 MS Australia 2012

Pfleger 2010 Tousignant 2011

Polman 2005 Trisolini 2010

Poser 1983 Van Tulder 2003

Rogante 2010 WHO 2001

Russell 2011 WHO 2008

Characteristics of included studies [ordered by study ID]

Study period: one month period of July 2010

Telerehabilitation for persons with multiple sclerosis (Review) 26

Exclusion: not specified

Interventions (similar to Motl 2011)

Outcomes Primary outcome: Physical activity: GLTEQ