Neuropsychological rehabilitation for multiple sclerosis.

Neuropsychological rehabilitation for multiple sclerosis (Review) Rosti-Otajärvi EM, Hämäläinen PI

This is a reprint of a Cochrane review, prepared and maintained by The Cochrane Collaboration and published in The Cochrane Library 2014, Issue 2 http://www.thecochranelibrary.com

Neuropsychological rehabilitation for multiple sclerosis (Review) Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.

TABLE OF CONTENTS HEADER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ABSTRACT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . PLAIN LANGUAGE SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . SUMMARY OF FINDINGS FOR THE MAIN COMPARISON . . . . . . . . . . . . . . . . . . . BACKGROUND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . OBJECTIVES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . METHODS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Figure 1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Figure 2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Figure 3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Figure 4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Figure 5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Figure 6. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ADDITIONAL SUMMARY OF FINDINGS . . . . . . . . . . . . . . . . . . . . . . . . . . DISCUSSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . AUTHORS’ CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ACKNOWLEDGEMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . CHARACTERISTICS OF STUDIES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . DATA AND ANALYSES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Analysis 1.1. Comparison 1 Cognitive training versus any control, Outcome 1 Attention. . . . . . . . . . . Analysis 1.2. Comparison 1 Cognitive training versus any control, Outcome 2 Information processing speed. . . . Analysis 1.3. Comparison 1 Cognitive training versus any control, Outcome 3 Memory span. . . . . . . . . Analysis 1.4. Comparison 1 Cognitive training versus any control, Outcome 4 Working memory. . . . . . . . Analysis 1.5. Comparison 1 Cognitive training versus any control, Outcome 5 Immediate verbal memory. . . . . Analysis 1.6. Comparison 1 Cognitive training versus any control, Outcome 6 Immediate visual memory. . . . . Analysis 1.7. Comparison 1 Cognitive training versus any control, Outcome 7 Delayed memory. . . . . . . . Analysis 1.8. Comparison 1 Cognitive training versus any control, Outcome 8 Executive functions. . . . . . . Analysis 1.9. Comparison 1 Cognitive training versus any control, Outcome 9 Verbal functions. . . . . . . . Analysis 1.10. Comparison 1 Cognitive training versus any control, Outcome 10 Depression. . . . . . . . . Analysis 1.11. Comparison 1 Cognitive training versus any control, Outcome 11 Quality of life. . . . . . . . Analysis 1.12. Comparison 1 Cognitive training versus any control, Outcome 12 Fatigue. . . . . . . . . . . Analysis 1.13. Comparison 1 Cognitive training versus any control, Outcome 13 Anxiety. . . . . . . . . . . Analysis 2.1. Comparison 2 Cognitive training versus any control (longitudinal follow-up), Outcome 1 Attention. . Analysis 2.2. Comparison 2 Cognitive training versus any control (longitudinal follow-up), Outcome 2 Immediate verbal memory. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Analysis 2.3. Comparison 2 Cognitive training versus any control (longitudinal follow-up), Outcome 3 Immediate visual memory. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Analysis 2.4. Comparison 2 Cognitive training versus any control (longitudinal follow-up), Outcome 4 Delayed memory. Analysis 2.5. Comparison 2 Cognitive training versus any control (longitudinal follow-up), Outcome 5 Verbal functions. Analysis 2.6. Comparison 2 Cognitive training versus any control (longitudinal follow-up), Outcome 6 Depression. . Analysis 2.7. Comparison 2 Cognitive training versus any control (longitudinal follow-up), Outcome 7 Quality of life. Analysis 3.1. Comparison 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control, Outcome 1 Attention. . . . . . . . . . . . . . . . . . . . . . . . . . . . . Analysis 3.2. Comparison 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control, Outcome 2 Information processing speed. . . . . . . . . . . . . . . . . . . . . . Analysis 3.3. Comparison 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control, Outcome 3 Memory span. . . . . . . . . . . . . . . . . . . . . . . . . . . Analysis 3.4. Comparison 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control, Outcome 4 Working memory. . . . . . . . . . . . . . . . . . . . . . . . . . Neuropsychological rehabilitation for multiple sclerosis (Review) Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.

1 1 2 3 6 7 7 9 10 13 14 16 18 19 20 24 26 27 27 34 71 80 82 83 84 85 87 88 89 90 91 92 93 94 95 96 96 97 98 98 99 100 102 103 104 i

Analysis 3.5. Comparison 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control, Outcome 5 Immediate verbal memory. . . . . . . . . . . . . . . . . . . . . . . Analysis 3.6. Comparison 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control, Outcome 6 Immediate visual memory. . . . . . . . . . . . . . . . . . . . . . . Analysis 3.7. Comparison 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control, Outcome 7 Delayed memory. . . . . . . . . . . . . . . . . . . . . . . . . . Analysis 3.8. Comparison 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control, Outcome 8 Executive functions. . . . . . . . . . . . . . . . . . . . . . . . . Analysis 3.9. Comparison 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control, Outcome 9 Visual functions. . . . . . . . . . . . . . . . . . . . . . . . . . Analysis 3.10. Comparison 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control, Outcome 10 Verbal functions. . . . . . . . . . . . . . . . . . . . . . . . . Analysis 3.11. Comparison 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control, Outcome 11 Everyday cognitive performance/patient’s report. . . . . . . . . . . . . . Analysis 3.12. Comparison 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control, Outcome 12 Everyday cognitive performance/carer’s report. . . . . . . . . . . . . . Analysis 3.13. Comparison 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control, Outcome 13 Depression. . . . . . . . . . . . . . . . . . . . . . . . . . Analysis 3.14. Comparison 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control, Outcome 14 Quality of life. . . . . . . . . . . . . . . . . . . . . . . . . Analysis 3.15. Comparison 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control, Outcome 15 Fatigue. . . . . . . . . . . . . . . . . . . . . . . . . . . . Analysis 3.16. Comparison 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control, Outcome 16 Anxiety. . . . . . . . . . . . . . . . . . . . . . . . . . . . Analysis 3.17. Comparison 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control, Outcome 17 Impact of the disease. . . . . . . . . . . . . . . . . . . . . . . Analysis 4.1. Comparison 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up), Outcome 1 Attention. . . . . . . . . . . . . . . . . . . . Analysis 4.2. Comparison 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up), Outcome 2 Information processing speed. . . . . . . . . . . . . Analysis 4.3. Comparison 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up), Outcome 3 Working memory. . . . . . . . . . . . . . . . . Analysis 4.4. Comparison 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up), Outcome 4 Immediate verbal memory. . . . . . . . . . . . . . Analysis 4.5. Comparison 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up), Outcome 5 Immediate visual memory. . . . . . . . . . . . . . Analysis 4.6. Comparison 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up), Outcome 6 Delayed memory. . . . . . . . . . . . . . . . . . Analysis 4.7. Comparison 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up), Outcome 7 Executive functions. . . . . . . . . . . . . . . . . Analysis 4.8. Comparison 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up), Outcome 8 Visual functions. . . . . . . . . . . . . . . . . . Analysis 4.9. Comparison 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up), Outcome 9 Verbal functions. . . . . . . . . . . . . . . . . . Analysis 4.10. Comparison 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up), Outcome 10 Everyday cognitive performance/patient’s report. . . . . Analysis 4.11. Comparison 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up), Outcome 11 Everyday cognitive performance/carer’s report. . . . . . Analysis 4.12. Comparison 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up), Outcome 12 Depression. . . . . . . . . . . . . . . . . . Analysis 4.13. Comparison 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up), Outcome 13 Quality of life. . . . . . . . . . . . . . . . . Neuropsychological rehabilitation for multiple sclerosis (Review) Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.

105 106 107 109 110 111 112 114 115 116 117 118 119 120 122 123 124 125 126 127 128 129 130 132 133 134 ii

Analysis 4.14. Comparison 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up), Outcome 14 Fatigue. . . . . . . . . . . . . . . . . . . Analysis 4.15. Comparison 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up), Outcome 15 Anxiety. . . . . . . . . . . . . . . . . . . Analysis 4.16. Comparison 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up), Outcome 16 Impact of the disease. . . . . . . . . . . . . . ADDITIONAL TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . APPENDICES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . WHAT’S NEW . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . CONTRIBUTIONS OF AUTHORS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . DECLARATIONS OF INTEREST . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . SOURCES OF SUPPORT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . INDEX TERMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .


135 136 137 137 140 143 144 144 144 144

iii

[Intervention Review]

Neuropsychological rehabilitation for multiple sclerosis Eija M Rosti-Otajärvi1 , Päivi I Hämäläinen2 1 Department

of Neurology and Rehabilitation, Tampere University Hospital, Tampere, Finland. 2 Masku neurological rehabilitation centre, Finnish MS Association, Masku, Finland Contact address: Päivi I Hämäläinen, Masku neurological rehabilitation centre, Finnish MS Association, Seppäläntie 90, Masku, 21250, Finland. [email protected]. Editorial group: Cochrane Multiple Sclerosis and Rare Diseases of the Central Nervous System Group. Publication status and date: New search for studies and content updated (conclusions changed), published in Issue 2, 2014. Review content assessed as up-to-date: 28 May 2013. Citation: Rosti-Otajärvi EM, Hämäläinen PI. Neuropsychological rehabilitation for multiple sclerosis. Cochrane Database of Systematic Reviews 2014, Issue 2. Art. No.: CD009131. DOI: 10.1002/14651858.CD009131.pub3. Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.

ABSTRACT Background This is an update of the Cochrane review ’Neuropsychological rehabilitation for multiple sclerosis’ (first published in The Cochrane Library 2011, Issue 11). Cognitive deficits are a common manifestation of multiple sclerosis (MS) and have a significant effect on the patient’s quality of life. Alleviation of the harmful effects caused by these deficits should be a major goal of MS research and practice. Objectives To assess the effects of neuropsychological/cognitive rehabilitation on health-related factors, such as cognitive performance and emotional well-being in patients with MS. Search methods The Cochrane Multiple Sclerosis and Rare Diseases of the Central Nervous System Group Trials Search Co-ordinator searched their Specialised Register which, among other sources, contains trials from CENTRAL (The Cochrane Library 2013, Issue 2), MEDLINE, EMBASE, CINAHL, LILACS, PEDro and clinical trials registries (28 May 2013). We contacted authors of the studies for additional information. Selection criteria Randomised controlled trials (RCTs) and quasi-randomised trials evaluating the effects of neuropsychological rehabilitation in MS compared to other interventions or no intervention. Data collection and analysis Two review authors individually judged the eligibility of the included studies, assessed risk of bias and extracted data. We combined results quantitatively in meta-analyses according to the intervention type: 1) cognitive training and 2) cognitive training combined with other neuropsychological rehabilitation methods. Neuropsychological rehabilitation for multiple sclerosis (Review) Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.

1

Main results Twenty studies (986 participants; 966 MS participants and 20 healthy controls) fulfilled the inclusion criteria. The mean age of the participants was 44.6 years, mean length of education was 12.3 years and 70% of the participants were women. Most of the participants had a relapsing-remitting course of disease. The mean Expanded Disability Status Scale score was 3.2 and the mean duration of disease was 14.0 years. On the basis of these studies, we found low-level evidence that neuropsychological rehabilitation reduces cognitive symptoms in MS. Cognitive training was found to improve memory span (standardised mean difference (SMD) 0.54, 95% confidence interval (CI) 0.20 to 0.88, P = 0.002) and working memory (SMD 0.33, 95% CI 0.09 to 0.57, P = 0.006). Cognitive training combined with other neuropsychological rehabilitation methods was found to improve attention (SMD 0.15, 95% CI 0.01 to 0.28, P = 0.03), immediate verbal memory (SMD 0.31, 95% CI 0.08 to 0.54, P = 0.008) and delayed memory (SMD 0.22, 95% CI 0.02 to 0.42, P = 0.03). There was no evidence of an effect of neuropsychological rehabilitation on emotional functions. The overall quality, as well as the comparability of the included studies, was relatively low due to methodological limitations and heterogeneity of interventions and outcome measures. Although most of the pooled results in the meta-analyses yielded no significant findings, 18 of the 20 studies showed some evidence of positive effects when the studies were individually analysed. Authors’ conclusions This review found low-level evidence for positive effects of neuropsychological rehabilitation in MS. The interventions and outcome measures included in the review were heterogeneous, which limited the comparability of the studies. New trials may therefore change the strength and direction of the evidence.

PLAIN LANGUAGE SUMMARY Rehabilitation for cognitive dysfunction in people with multiple sclerosis This is an update of the Cochrane review ’Neuropsychological rehabilitation for multiple sclerosis’ (first published in The Cochrane Library 2011, Issue 11). Demyelinating disease is any condition that results in damage to the protecting covering (myelin sheath) that surrounds nerve fibers in brain and spinal cord. Multiple sclerosis (MS) has used to be regarded simply as a demyelinating disease, however recent research has shown that widespread damage to neurons (nerve cells) and grey matter changes are central features of MS. This has emphasised cognitive dysfunction, like deficits in memory or attention, as being one of the major symptoms of the disease. Cognitive deficits are common in MS, occurring in about 50% to 60% of patients. These deficits can have a multidimensional impact on patients’ activities of daily living and should be taken into account in their treatment and rehabilitation. Neuropsychological rehabilitation aims to 1) reduce cognitive deficits, 2) reduce the harmful effects of cognitive impairments and 3) support patients’ awareness and ability to take cognitive impairments into account in daily living. The aim of this review was to evaluate the effects of cognitive (neuropsychological) rehabilitation in MS. We did this by considering the effects of rehabilitation on cognitive test performance and everyday cognitive performance, as well as on depression, fatigue, personality/ behaviour disturbances, anxiety and quality of life. Twenty relevant studies comprising a total of 986 participants (966 MS participants and 20 healthy controls) were identified and included in this review. Low-level evidence was found that neuropsychological rehabilitation reduces cognitive symptoms in MS. However, when analysed individually, 18 out of the 20 studies showed positive effects. Cognitive training was found to improve memory span and working memory. Cognitive training combined with other neuropsychological rehabilitation methods was found to improve attention, immediate verbal memory and delayed memory. It is worth noting that the small numbers of patients in the studies and some methodological weaknesses reduce the level of the evidence. To further strengthen the evidence well-designed, high-quality studies are needed.


2


S U M M A R Y O F F I N D I N G S F O R T H E M A I N C O M P A R I S O N [Explanation]

Cognitive training versus any control for multiple sclerosis Patient or population: patients with multiple sclerosis Settings: outpatient Intervention: cognitive training versus any control Outcomes

Attention

Illustrative comparative risks* (95% CI)

Assumed risk

Corresponding risk

Control

Cognitive training versus any control

No of participants (studies)

The mean attention score The mean attention score SMD 0.06 (95% CI -0.1 573 in the control groups was in the intervention groups to 0.23), (5 studies) 12.6 was P = 0.46 0.06 standard deviations higher (0.1 lower to 0.23 higher)

Information processing The mean information speed processing speed score in the control groups was -410.9

Memory span

Relative effect (95% CI)

The mean information SMD 0.15 (95% CI -0.33 176 processing speed score to 0.62), (4 studies) in the intervention groups P = 0.55 was 0.15 standard deviations higher (0.33 lower to 0.62 higher)

3

The mean memory span The mean memory span SMD 0.54 (95% CI 0.2 to 150 score in the control score in the intervention 0.88), (2 studies) groups was 5.8 groups was P = 0.002 0.54 standard deviations higher (0.2 to 0.88 higher)

Quality of the evidence (GRADE)

Comment

⊕⊕

low1

⊕

very low1,2,3

⊕⊕

low3,4

Time-related scores were entered as negative variables.


Working memory

The mean working mem- The mean working mem- SMD 0.33 (95% CI 0.09 288 ory score in the control ory score in the interven- to 0.57), (3 studies) groups was 8.6 tion groups was P = 0.006 0.33 standard deviations higher (0.09 to 0.57 higher)

⊕

very low1,3

Immediate verbal mem- The mean immediate ver- The mean immediate ver- SMD 0.2 (95% CI -0.02 360 ory bal memory score in the bal memory score in the to 0.41), (4 studies) control groups was 4.3 intervention groups was P = 0.08 0.2 standard deviations higher (0.02 lower to 0.41 higher)

⊕⊕

low3,4

Executive functions

The mean executive func- The mean executive func- SMD 0.35 (95% CI -0.03 112 tions score in the control tions score in the inter- to 0.73), (2 studies) groups was -6.6 vention groups was P = 0.07 0.35 standard deviations higher (0.03 lower to 0.73 higher)

⊕

very low1,3

Error scores were entered as negative variables.

Depression

The mean depression The mean depression SMD 0.26 (95% CI -0.23 196 score in the control score in the intervention to 0.75), (5 studies) groups was -28.0 groups was P = 0.29 0.26 standard deviations higher (0.23 lower to 0.75 higher)

⊕

very low1,2,3

Depression scores were entered as negative variables.

*The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; SMD: standardised mean difference

4


GRADE Working Group grades of evidence High quality: Further research is very unlikely to change our confidence in the estimate of effect. 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. 1 Several

crucial limitations in the implementation of the study. unexplained heterogeneity. 3 Low number of participants. 4 One crucial limitation in the implementation of the study. 2 High,

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

5

Description of the intervention BACKGROUND

Description of the condition Multiple sclerosis (MS) is a chronic inflammatory disease of the central nervous system (CNS) causing damage to both white and grey matter. The average age at disease onset is 30 years (Barten 2010). MS is the most common disabling neurological disease in young and middle-aged adults, and the economic effects of the disease are noticeable (Barten 2010). MS is characterised by a marked variability in its clinical symptoms and their course, while the unpredictable nature of the disease trajectory increases the distress it causes patients. Cognitive deficits are a common manifestation of MS, occurring in about 50% to 60% of patients (Amato 2006). The neuropathological background of the deficits is considered to relate to the variety of white and grey matter damage (Filippi 2010). There is no direct relationship between cognitive deficits and physical disability, disease duration or course of the disease (Amato 2008). On the other hand, it has been found that high primary cognitive capacity may protect cognitive functions by creating ’cognitive reserves’, thus contributing to the development of functional reorganisation processes (Bonnet 2006; Sumowski 2009). MS-related cognitive dysfunction is heterogeneous and its evolution is highly individual among patients. In longitudinal studies, both stable cognitive status (Jennekens-Schinkel 1990; Kujala 1997; Mariani 1991; Rosti 2007; Sperling 2001) and decline (Amato 2001; Feinstein 1992; Kujala 1997; Zivadinov 2001) have been reported. Memory and learning dysfunction and slowed information processing speed are regarded as the most common MSrelated cognitive deficits (Calabrese 2006; Rogers 2007). Slowed information processing speed, in particular, is thought to underlie and be the core feature of cognitive problems observed in MS (DeLuca 2004; Henry 2006) and can negatively affect other aspects of cognitive function (Calabrese 2006; Feinstein 2007). Deficits in complex attention and executive functions, in visual perception and in language (verbal fluency and naming) can also occur (Fischer 2001). The functional consequences of MS-related cognitive impairment can be striking and have a multidimensional impact on patients’ activities of daily living. Thus, MS-related cognitive impairment and its recovery should be taken into account in the patient’s treatment and rehabilitation regimen. Indeed, cognitive deficits may have both physical and mental effects, leading to symptoms of cognitive fatigue (DeLuca 2005), and may have an impact on (physical) independence, employment, social and recreational activities (Chiaravalloti 2008), driving skills (Kotterba 2003; Lincoln 2008) and on physical rehabilitation outcome (Langdon 1999), as well as caregiver strain (Khan 2007).

Neuropsychological rehabilitation is based on a comprehensive neuropsychological assessment where the patient’s individual cognitive strengths and weaknesses are evaluated. In order to reduce the harmful effects on the patient’s everyday life, neuropsychological rehabilitation involves teaching compensatory strategies, using aids, supporting the patient’s awareness and counselling patients and their significant others in how to cope better with cognitive problems. While psychological interventions aim to treat mood disorders such as depression or anxiety, reduce stress, enhance self efficacy and self esteem, and improve psychological coping skills, neuropsychological interventions focus on the improvement of cognitive skills, by supporting the patient’s awareness of their own cognitive strengths and weaknesses. Thus, neuropsychological rehabilitation is provided by a specialised clinical neuropsychologist.

How the intervention might work Neuropsychological rehabilitation aims to: 1. reduce cognitive deficits; 2. reduce the harmful effects of cognitive impairments; 3. support the patients’ awareness of their cognitive impairment to enable them to cope in everyday living. Functional magnetic resonance imaging (fMRI) studies on cognitive function in MS have revealed that functional brain reorganisation often takes place spontaneously to compensate for tissue damage (Penner 2007). Functional brain reorganisation after a brain injury mainly consists of an increase in the extent of brain areas activated, even recruiting additional areas (Pantano 2006). Neuroplasticity may reduce the clinical expression of cognitive symptoms and it has been speculated that cognitive rehabilitation might be seen as a procedure to enhance plasticity processes in the brain (Chiaravalloti 2012; Filippi 2012; Leavitt 2012; Parisi 2012; Penner 2007). The scientific literature reports some evidence of cognitive rehabilitation reducing cognitive impairment. Attention and memory retraining has been shown to be effective after traumatic brain injury, as has language and perception training in persons with stroke (Cicerone 2000). Furthermore, it has been suggested that by providing patients with knowledge about their cognitive strengths and weaknesses, it is possible to increase their awareness and ability to find ways to cope with persisting neuropsychological deficits (Wilson 2008). On the other hand, teaching compensatory strategies may help patients to cope with their cognitive impairments (Wilson 2008). Finally, in order to help patients to understand and accept their cognitive problems and increase their psychological well-being, neuropsychological support in individual and group rehabilitation settings appears to be effective (Mateer 2005).

Why it is important to do this review


6

The effects of neuropsychological rehabilitation in MS have not been as widely studied as they have, for example, in stroke patients, where the advantages of neuropsychological rehabilitation have been more extensively evaluated (Bowen 2007; das Nair 2007; Lincoln 2000). The heterogeneous and progressive nature of MSrelated cognitive deficits, in addition to the difficulty in investigating slow information processing, which appears to be the main cognitive impairment in the MS population, may explain the reluctance of researchers to conduct this type of study. In the Cochrane review by Thomas et al (Thomas 2006), the effectiveness of psychological interventions for MS was evaluated, based on the evidence from 16 randomised controlled trials (RCTs). However, only five of these trials focused on neuropsychological interventions. This review concentrated on evaluating whether psychological interventions had an effect on quality of life, psychiatric symptoms (e.g. depression and anxiety), psychological functioning (e.g. emotions, self efficacy, self esteem) and neurological disability, and as only one component of cognitive function. In addition to Thomas 2006 another a Cochrane review focused solely on memory rehabilitation in MS (das Nair 2012a) and did not cover other cognitive functions or neuropsychological rehabilitation. Our review, on the other hand, will focus on evaluating whether neuropsychological rehabilitation has an effect on improving cognitive performance and patients’ ability to cope with their cognitive deficits. It is important to undertake this review because cognitive deficits are common in MS and may have a multidimensional impact on patients’ quality of life. Medication cannot prevent cognitive deficits or their progression and simple psychological interventions do not attain the same benefits. This systematic review will evaluate the published literature on this topic and assess its quality. This is an update of the Cochrane review ’Neuropsychological rehabilitation for multiple sclerosis’ (first published in The Cochrane Library 2011, Issue 11).

We defined quasi-randomised allocation as the use of factors such as date of admission, hospital number, alternation, date of birth or assignment to comparable groups with respect to clinical and demographic factors. Quasi-randomised trials were included because of their number in the scientific literature; their exclusion would have meant discarding most of the available data. Crossover studies were not included. Types of participants Patients with a clinically definite MS diagnosis according to the Poser (Poser 1983) or McDonald (McDonald 2001; Polman 2005) criteria. We included studies in the review regardless of the clinical course of the disease, or its severity or duration. Studies concerning both MS patients and other patient populations were also included only when the results for the MS patients could be separated from the other clinical population. Types of interventions Neuropsychological/cognitive interventions compared with a control condition were included. Studies with neuropsychological/ cognitive rehabilitation as a part of multidisciplinary rehabilitation were also included when the share of the neuropsychological rehabilitation in the intervention was clearly reported. All neuropsychological rehabilitation approaches were included, such as cognitive retraining, teaching of compensatory strategies and use of aids, and neuropsychological counselling or support aimed at reducing the harmful effects of cognitive/behavioural problems (see section Description of the intervention). The target of the intervention had to be the reduction of the impairments caused by or related to the neuropsychological deficits, such as memory or attention disorders. We set no restrictions on delivery, duration, intensity or co-interventions for interventions or comparators. Psychological interventions and drug studies were excluded. We set no other criteria for the control group intervention. Types of outcome measures

OBJECTIVES To assess the effects of neuropsychological/cognitive rehabilitation on health-related factors, such as cognitive performance and emotional well-being in patients with MS.

We classified assessments according to whether they were observed: 1. immediately post-treatment (within a month posttreatment); 2. at longer-term follow-up (second measurement more than one month post-treatment).

METHODS

Primary outcomes

Criteria for considering studies for this review

• Measures of cognitive function: we selected standardised neuropsychological tests to obtain an objective and worldrecognised score, in addition to self reported questionnaires to assess the perceived impact of cognitive dysfunction on everyday life

Types of studies Randomised controlled trials (RCTs) and quasi-randomised trials.

Outcome measures had to be standardised, objective research tools for measuring cognitive performance, such as the Wechsler Adult


7

Intelligence Scale (WAIS, Wechsler 1997a), the Wechsler Memory Scale (WMS, Wechsler 1997b), the Paced Auditory Serial Addition Test (PASAT, Gronwall 1977), the Symbol Digit Modalities Test (SDMT, Smith 1973) etc. For evaluating everyday cognitive performance we selected subjective self reports and carer ratings for measuring the harmful effects of cognitive problems and/or meta-cognition experienced, such as the Multiple Sclerosis Neuropsychological Questionnaire (Benedict 2003), the Memory Functioning Questionnaire (MFQ, Gilewski 1990), the Everyday Memory Questionnaire (EMQ, Sunderland 1983), the Dysexecutive Syndrome Questionnaire (DEX, Wilson 1996) etc.

The keywords used to search for this review are listed in Appendix 1. In addition, we performed an expanded search to identify all quasi-randomised clinical trial articles in the following databases: PubMed (Appendix 2); EMBASE (Appendix 3) and CINAHL (Appendix 4), up to 28 May 2013. Searching other resources In addition to the electronic searches, we examined the reference lists of the included studies. We also used personal contact with experts in the field to identify the studies.

Secondary outcomes

Secondary outcomes were measures of: • depression (such as the Beck Depression Inventory (BDI, Beck 1961)); • fatigue (such as the Fatigue Scale for Motor and Cognitive Functions (FSMC, Penner 2005)); • personality disturbance (such as the NEO Personality Inventory (NEO-PI, Hogan 1969)); • anxiety (such as the State-Trait Anxiety Inventory (STAI, Spielberger 1983)); and • quality of life (such as the SF-36 (Ware 1992)).

Data collection and analysis

As secondary outcome measures, we also considered non-standardised measures such as return to work or activity level.

Two review authors (ER and PH) extracted data from the studies. The first author (ER) tabulated the data, which were checked and completed by the second author (PH). We reached a consensus for every item included in the final tables (Characteristics of included studies; Data and analyses). We extracted and tabulated data concerning: 1. participants (number, drop-outs, age, sex, years of education, MS disease course, severity and duration of the disease, exclusion criteria, research setting); 2. interventions delivered (type, cognitive domain targeted, number/frequency/duration of sessions and follow-up times); 3. results reported (primary and secondary outcome measures, timing of the measurements, main and secondary results).

Search methods for identification of studies We conducted a systematic search without language or date restrictions to identify all relevant published and unpublished randomised controlled trials and quasi-randomised clinical trials. Electronic searches The Cochrane Multiple Sclerosis and Rare Diseases of the Central Nervous System Trials Search Co-ordinator searched their Specialised Register (28 May 2013) which, among other sources, contains: 1. Cochrane Central Register of Controlled Trials (CENTRAL) (2013, Issue 2); 2. MEDLINE (PubMed) (1966 to 28 May 2013); 3. EMBASE (Embase.com) (1974 to 28 May 2013); 4. CINAHL (EBSCO host) (1981 to 28 May 2013); 5. LILACS (Bireme) (1982 to 28 May 2013). In addition, we searched clinical trials registries: Clinicaltrials.gov ( www.clinicaltrials.gov) and the WHO International Clinical Trials Registry Portal (http://apps.who.int/trialsearch/). Information on the Specialised Register and details of the search strategies used to identify trials can be found in the ’Specialised Register’ section within the Cochrane Multiple Sclerosis and Rare Diseases of the Central Nervous System Group’s module.

Selection of studies Two review authors (ER and PH) independently selected articles and assessed study relevance in line with the predetermined inclusion criteria. Disagreements were discussed and resolved by consensus among the review authors. Data extraction and management

Assessment of risk of bias in included studies Two review authors (ER and PH) independently graded the selected studies using a simple contingency form following the domain-based evaluation described in the Cochrane Handbook for Systematic Reviews of Interventions 5.1.0 (Higgins 2011). The authors compared evaluations and discussed and resolved any disagreements. The authors assessed the following domains as ’low risk of bias’, ’unclear risk of bias’ or ’high risk of bias’: 1. random sequence generation (adequate use of a random component in the sequence generation); 2. allocation concealment (adequate use of allocation concealment so that participants or investigators enrolling participants could not foresee assignment);


8

3. blinding of participants, personnel and outcome assessors (knowledge of the allocated interventions was adequately prevented during the study); 4. incomplete outcome data (acceptable drop-out rate (< 20%)); 5. selective reporting (all intended outcome data presented for intervention and control groups); and 6. other bias (e.g. similarity in groups’ baseline characteristics regarding the most important prognostic indicators; acceptable compliance with intervention). We reported the results of the ’Risk of bias’ assessments for each individual study in the ’Risk of bias’ tables and presented the overall quality of the findings in Summary of findings for the main comparison and Summary of findings 2. We used the GRADE software to provide an overall grading of the quality of the evidence (GRADE 2008). Measures of treatment effect We analysed data using RevMan 5.2 (RevMan 2012). The included studies evaluated different cognitive functions using various tests and rating scales. We grouped the cognitive tests employed in the included studies into cognitive domains according to the functions they are traditionally used to measure (Lezak 2004) and we evaluated the effects of neuropsychological rehabilitation separately for those domains. For data where a higher score denotes a worse performance (time-related and error variables in cognitive tests, outcome measures for depression, fatigue and anxiety) we entered the mean at the post-intervention time point as a negative variable. The data for continuous outcomes were the mean and standard deviation (SD) of each post-intervention outcome variable and the number of participants for each treatment group at each assessment point. We entered test scores at the end of treatment and during further follow-ups in the analyses. We analysed treatment effects using mean differences (MD) for continuous outcome variables. We calculated the standardised mean difference (SMD) when continuous outcome variables were measured with different scales. For all treatment effects, we provided 95% confidence intervals (CI).

Assessment of heterogeneity We employed the I2 statistic to evaluate heterogeneity of study variables statistically. I2 values of 25%, 50% and 75% corresponded to low, medium and high levels of heterogeneity, respectively. When the I2 statistic was significant (at ≥ 50%), we calculated estimates of treatment effects using a random-effects model (which provides broader confidence intervals than a fixed-effect model). When heterogeneity was not found, we adopted a fixed-effect model. Assessment of reporting biases To assess possible publication bias, we created funnel plots for those outcomes with at least 10 variables. Data synthesis To provide a meaningful summary, two authors (ER and PH) considered whether the clinical and methodological characteristics of the included studies were similar enough for meta-analysis. We combined results quantitatively in meta-analyses according to the intervention type: 1. cognitive training; and 2. cognitive training combined with other neuropsychological rehabilitation methods. Additionally, we used descriptive analysis in reporting the results of multimodal neuropsychological rehabilitation. Sensitivity analysis We carried out sensitivity analyses to assess the robustness of the results by: 1. excluding studies with inadequate concealment of allocation; 2. excluding studies in which outcome evaluation was not blinded or blinding remained unclear; 3. excluding studies in which loss to follow-up remained unclear or was greater than 20%.

RESULTS Unit of analysis issues We included studies with multiple intervention and control groups, which we analysed by pooling together the data from all the intervention groups into a single intervention group and from all the control groups into a single control group.

See the Characteristics of included studies and Characteristics of excluded studies tables.

Dealing with missing data

Results of the search

When data were not available or were unclear in the original study reports, we contacted the authors of the studies in question in order to obtain more precise information.

The literature search process is described in Figure 1. Overall, we retrieved 6500 studies (including overlapping search results from different databases). From these, 20 studies were included. Three

Description of studies


9

studies (Filippi 2012; Mattioli 2012a; Vogt 2009) were related to other included studies. Both authors of the review acted as authors in one of the evaluated studies (Mäntynen 2013). The journal titles, names of the authors and institutions, sources of funding and results were available to the review authors in all phases of the review. Figure 1. Flow diagram of the study selection process. Filippi 2012, Mattioli 2012a, and Vogt 2009 were related studies, so 17 were included in analyses and 3 were related studies (not included in analyses).

Included studies The 20 studies included were published between 1993 and 2013 and conducted in 10 different countries: six in Italy, four in the United States, two in Switzerland, two in Germany and one in Denmark, Great Britain, Austria, Israel, France and Finland. One study was written in German and the others in English. All except

one (Mäntynen 2013) were single-centre studies. We obtained further information regarding the randomisation method for three studies (Benedict 2000; Mattioli 2010; Tesar 2005) through email contact with study authors. We obtained statistical data for outcome measures for four studies (Chiaravalloti 2005; Lincoln 2002: data for meta-cognition outcomes and subjective experience and carers’ reports of the harmful effects of cog-


10

nitive problems; Mattioli 2010; Shatil 2010: data for depression and fatigue). For Jonsson 1993 the original data were no longer available. Participants

The 20 trials included 986 participants: 966 MS patients and 20 healthy controls. Detailed information about the participants is given in the Characteristics of included studies table and the participants’ characteristics are summarised in Table 1. The number of participants in the studies varied from 15 to 240 (mean 58) and the drop-out rate ranged from 0% to 57% (mean 9%; drop-out rates were not reported in Hildebrandt 2007). The proportion of women among the participants varied from 48% to 100% (mean 70%; gender distribution was not reported in Brissart 2012). The mean age of the participants varied from 32.7 to 60.0 years (mean 44.6) and mean length of education ranged from 8.5 to 14.8 years (mean 12.3; education years were not reported in Fink 2010, Mendoza 2001 or Tesar 2005). Most patients (414; 43%) had a relapsing-remitting course of disease (disease course or number of participants with a certain disease course were not reported in Mendoza 2001, Mendozzi 1998, Shatil 2010 or Stuifbergen 2012). The mean Expanded Disability Status Scale (EDSS) scores varied from 1.4 to 5.6 (mean 3.2) among patients, and the mean duration of disease from 4.7 to 17.5 years (mean 14.0; EDSS scores were not reported in Chiaravalloti 2005, Fink 2010, Lincoln 2002 or Mendoza 2001 and duration of disease in Benedict 2000, Lincoln 2002, Mendoza 2001 or Shatil 2010). In 13 of the studies (65%), cognitive impairment, either self reported, objectively demonstrated or both, was one of the inclusion criteria. The heterogeneity of the methods used to define and evaluate cognitive impairment did not make it possible to characterise the overall degree of cognitive impairment in the study population. Interventions

Detailed information about interventions is presented in the Characteristics of included studies table and the intervention characteristics are summarised in Table 2. Cognitive training was implemented in 18 studies (90% of the studies; Brissart 2012; Cerasa 2012; Chiaravalloti 2005; Filippi 2012; Fink 2010; Hildebrandt 2007; Jonsson 1993; Lincoln 2002; Mattioli 2010; Mattioli 2012a; Mendozzi 1998; Mäntynen 2013; Shatil 2010; Solari 2004; Stuifbergen 2012; Tesar 2005; Vogt 2008; Vogt 2009). In nine out of the 18 studies (50%) (Brissart 2012; Chiaravalloti 2005; Fink 2010; Hildebrandt 2007; Jonsson 1993; Lincoln 2002; Mäntynen 2013; Tesar 2005; Stuifbergen 2012) the intervention programme consisted of learning compensatory strategies in addition to cognitive training. In 15 studies (83%), cognitive training was computer-assisted and in three studies it was performed

as paper-pencil tasks. Computer-assisted training was conducted at home in six (40%) and in clinic in nine (60%) studies. The computer-assisted programs were: RehaCom (n = 7; Cerasa 2012; Filippi 2012; Mattioli 2010; Mattioli 2012a; Mendozzi 1998; Solari 2004; Tesar 2005), BrainStim (n = 2; Vogt 2008; Vogt 2009), Vilat-G 1.0 (n = 1; Hildebrandt 2007), ProCogSEP (n = 1; Brissart 2012), Foramen Rehab (n = 1 Mäntynen 2013) and an internet-based program (www.neuropsychonline.com) (n = 1 Stuifbergen 2012). Cognitive training was described according to five cognitive domains: • Memory training: memorising and recalling of word lists, pictures and figures, routes, numbers and stories. • Attentional training: turned, inversed and reflected texts, ’two in one’ pictures, labyrinths, searching of figures and train driving with the monitoring of distractors. • Visuospatial training: mosaic games and navigating tasks. • Training of executive functions: organising, planning and developing solution strategies employing realistic simulations of a set of scheduled dates and duties. • Problem-solving skills training: puzzles requiring deductive reasoning, organisation and analysis of facts. Learning of compensatory strategies was included in 11 of the studies (55%) (Benedict 2000; Brissart 2012; Chiaravalloti 2005; Fink 2010; Hildebrandt 2007; Jonsson 1993; Lincoln 2002; Mendoza 2001; Mäntynen 2013; Tesar 2005; Stuifbergen 2012) and mostly used for memory functions. Both external (calendar, diaries, notebooks, and lists) and internal (visualisation, logical content like story building, semantic categorisation and associations) memory strategies were used. Calendars and notebooks were also used in a wider context; for example calendars were used for planning everyday activities and notebooks for recording current needs. Building routines into behaviour and self control techniques were mentioned as compensatory strategies for attentional deficits, as were simplifying activities and visualisation for problem-solving and planning deficits. Furthermore, cognitivebehavioural and neuropsychotherapeutic methods were used to process cognitive and behavioural problems and to improve their management. Two out of 20 interventions (10%) did not include cognitive training. The methods employed in these studies were a cognitive-behavioural intervention to reduce behavioural problems (Benedict 2000) and the use of notebooks to intensify the observation of institutionalised patients’ needs (Mendoza 2001). The target of the intervention was a single cognitive function in seven studies (35%) (Cerasa 2012; Chiaravalloti 2005; Fink 2010; Mäntynen 2013; Mendozzi 1998; Vogt 2008; Vogt 2009). The primary target was usually memory function, either alone or together with other functions. In 12 studies (60%) (Brissart 2012; Filippi 2012; Hildebrandt 2007; Jonsson 1993; Lincoln 2002; Mattioli 2010; Mattioli 2012a; Mendoza 2001; Shatil 2010; Solari 2004; Stuifbergen 2012; Tesar 2005) the target was several cog-


11

nitive functions, mainly tailored to the patient’s individual symptoms, while in Benedict 2000 the target of the intervention was behaviour regulation. The reported number of intervention sessions in the studies varied from eight to 36 (mean 17) and the duration of each session from 25 minutes to two hours (mean one hour; sessions and their duration were not reported in all studies). The frequency of interventions varied from five times a week to twice per month (mean two times per week; frequency was not reported in all studies). The duration of the interventions varied from four weeks to six months (mean 9.5 weeks) and the entire follow-up time from immediate follow-up to one year (mean 25.3 weeks). Of the research settings, 15 were comparative studies with two groups (75%). In three studies (Lincoln 2002; Mendozzi 1998; Vogt 2009) there were three study groups and in one study (Vogt 2008) four (intervention and control groups of MS patients and healthy controls). The control group received no intervention in 13 of the studies (Filippi 2012; Fink 2010; Hildebrandt 2007; Lincoln 2002; Mattioli 2010; Mattioli 2012a; Mendoza 2001; Mäntynen 2013; Shatil 2010; Stuifbergen 2012; Tesar 2005; Vogt 2008; Vogt 2009) and unspecific cognitive training in five of the studies (Cerasa 2012; Chiaravalloti 2005; Jonsson 1993; Mendozzi 1998; Solari 2004). Other control interventions were: supportive psychotherapy (n = 1; Benedict 2000), neuropsychological evaluation with feedback (n = 1; Lincoln 2002), discussion program (n = 1; Brissart 2012), an intervention programme for healthy controls (n = 1; Vogt 2008) and low-intensity distributed training (n = 1; Vogt 2009).

Outcomes

Cognitive tests were used as baseline measurements in all 20 studies. Detailed information about the outcome measures is given in the Characteristics of included studies table. Cognitive tests were used as the outcome measures in most of the studies (90%, n = 18), except for Benedict 2000 and Lincoln 2002 where neuropsychological assessment was used only to document the similarity of the compared groups at the beginning of the study. Cognitive test results acted as the sole outcome measures only in two studies (Brissart 2012; Mendozzi 1998). Instead, it was more common that both cognitive tests and self rating questionnaires were used as outcome measures.

In most of the studies (70%) outcome was evaluated with self rating mood questionnaires (n = 14; Benedict 2000; Cerasa 2012; Chiaravalloti 2005; Hildebrandt 2007; Jonsson 1993; Mäntynen 2013; Mattioli 2010; Mattioli 2012a; Mendoza 2001; Shatil 2010; Solari 2004; Tesar 2005; Vogt 2008; Vogt 2009), of which the Beck Depression Inventory (BDI; Beck 1961) was the most commonly used (n = 8; Benedict 2000; Cerasa 2012; Chiaravalloti 2005; Hildebrandt 2007; Jonsson 1993; Mäntynen 2013; Mendoza 2001; Tesar 2005). In addition, other self report questionnaires were used, such as questionnaires on fatigue (n = 7; Cerasa 2012; Hildebrandt 2007; Mäntynen 2013; Shatil 2010; Tesar 2005; Vogt 2008; Vogt 2009), quality of life (n = 6; Hildebrandt 2007; Lincoln 2002; Mäntynen 2013; Mattioli 2010; Mattioli 2012a; Solari 2004), meta-cognition/subjectively experienced harmful effects of cognitive problems (n = 4; Chiaravalloti 2005; Lincoln 2002; Mäntynen 2013; Stuifbergen 2012), anxiety (n = 3; Cerasa 2012; Chiaravalloti 2005; Jonsson 1993), personality (n = 1; Benedict 2000) and behaviour disturbances (n = 1; Benedict 2000). In three studies (Benedict 2000; Lincoln 2002; Mäntynen 2013) a caregiver evaluated the patient’s social behaviour and personality, everyday memory and executive functions, as well as the caregiver’s own mood. The cognitive tests and questionnaires used varied among the studies. The four most commonly used outcome measures in the original studies are specified in Table 3. Only in three studies (Fink 2010; Mäntynen 2013; Solari 2004) were the primary outcome measures determined beforehand and separated from the secondary outcome measures. In 11 out of 20 studies (55%), the outcome was measured immediately post-treatment (within a month post-treatment) without longitudinal follow-up, while in the remaining nine studies (45%)(Chiaravalloti 2005; Fink 2010; Jonsson 1993; Lincoln 2002; Mattioli 2012a; Mäntynen 2013; Solari 2004; Stuifbergen 2012; Tesar 2005) longitudinal follow-ups were employed and varied from 11 weeks to one year post-treatment (mean 25.3 weeks). Excluded studies Based on the full text, 58 studies were excluded. Reasons for exclusion are reported in the Characteristics of excluded studies table.

Risk of bias in included studies See the ’Risk of bias’ graph (Figure 2) and ’Risk of bias’ summary (Figure 3).


12

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


13

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


14

Allocation The randomisation was adequate and unpredictable only in 35% of studies (n = 7) (Benedict 2000; Cerasa 2012; Jonsson 1993; Lincoln 2002; Mäntynen 2013; Solari 2004; Stuifbergen 2012). In these studies random number tables and a sealed opaque envelope system were used as randomisation and allocation concealment methods. In six of these studies (Cerasa 2012; Jonsson 1993; Lincoln 2002; Mäntynen 2013; Solari 2004; Stuifbergen 2012) the personnel who performed the randomisation process were not involved in the study itself. However, in most of the studies (65%), the randomisation procedure was inadequate and predictable (n = 13) (Brissart 2012; Chiaravalloti 2005; Fink 2010; Filippi 2012; Hildebrandt 2007; Mattioli 2010; Mattioli 2012a; Mendoza 2001; Mendozzi 1998; Shatil 2010; Tesar 2005; Vogt 2008; Vogt 2009); non-random components such as alternation, assignment to comparable groups with respect to clinical and demographic factors, and date of admission were used in the sequence generation processes. Blinding In most of the studies, the blinding of participants and personnel was not possible because of the characteristics of the interventions. In 74% of studies the blinding of the participant was not successful or it remained unclear whether it had succeeded in practice (n = 15) (Benedict 2000; Filippi 2012; Hildebrandt 2007; Jonsson 1993; Mäntynen 2013; Lincoln 2002; Mattioli 2010; Mattioli 2012a; Mendoza 2001; Mendozzi 1998; Shatil 2010; Stuifbergen 2012; Tesar 2005; Vogt 2008; Vogt 2009). The blinding of personnel was not successful or it was reported insufficiently in most of the studies (n = 16) (Benedict 2000; Brissart 2012; Cerasa 2012; Chiaravalloti 2005; Filippi 2012; Fink 2010; Jonsson 1993; Lincoln 2002; Mattioli 2010; Mattioli 2012a; Mendoza 2001; Mendozzi 1998; Mäntynen 2013; Solari 2004; Stuifbergen 2012; Tesar 2005). The blinding of personnel was judged to be successful only in those studies where training was carried out by the participant him/herself with a computeraided program (Hildebrandt 2007; Shatil 2010; Vogt 2008; Vogt 2009). The blinding of outcome assessors was slightly more successful, but it remained unclear or insufficient in 30% of the studies (n = 6) (Jonsson 1993; Mendoza 2001; Shatil 2010; Tesar 2005; Vogt 2008; Vogt 2009). However, only three studies (Jonsson 1993; Mäntynen 2013; Solari 2004) reported how the blinding of the assessor had succeeded in practice.

Vogt 2008; Vogt 2009). In nine studies (45%), the percentage of drop-outs and withdrawals was less than 20% (Cerasa 2012; Chiaravalloti 2005; Lincoln 2002; Mendoza 2001; Mendozzi 1998; Mäntynen 2013; Solari 2004; Stuifbergen 2012; Tesar 2005). In four studies (20%) the percentage of drop-outs and withdrawals exceeded 20% or was reported insufficiently (Fink 2010; Hildebrandt 2007; Jonsson 1993; Shatil 2010).

Selective reporting In most of the studies (80%) we did not find indications of selective reporting: all intended outcome data were present for the intervention and control groups during the intervention and followup phases (n = 16) (Benedict 2000; Brissart 2012; Cerasa 2012; Chiaravalloti 2005; Filippi 2012; Hildebrandt 2007; Mattioli 2010; Mattioli 2012a; Mendozzi 1998; Mäntynen 2013; Shatil 2010; Solari 2004; Stuifbergen 2012; Tesar 2005; Vogt 2008; Vogt 2009). In 10% of studies, only some of the pre-specified outcome measures were reported (n = 2) (Fink 2010; Jonsson 1993). In one study (Lincoln 2002) cognitive tests were not included as outcome measures although the intervention was cognitive training and the target was cognitive rehabilitation. Consequently, we judged that results for some key outcomes were not reported. Moreover, in one study data for all outcome measures (cognitive tests) were not reported (Mendoza 2001).

Other potential sources of bias We did not find any other sources of bias in 65% of the studies (n = 13) (Benedict 2000; Brissart 2012; Cerasa 2012; Filippi 2012; Jonsson 1993; Mattioli 2010; Mattioli 2012a; Mendozzi 1998; Mäntynen 2013; Tesar 2005; Stuifbergen 2012; Vogt 2008; Vogt 2009). However, in six studies (30%) the intervention and control groups were not similar at baseline or their similarity remained unclear for the most important prognostic indicators. Groups differed in demographic factors, such as age (Chiaravalloti 2005; Hildebrandt 2007; Mendoza 2001; Shatil 2010; Solari 2004) and gender (Hildebrandt 2007), or baseline characteristics were not reported in sufficient detail (Brissart 2012; Fink 2010). Furthermore, in one study compliance with the intervention was weak; only 46% of participants in the intervention group participated in the intervention as planned and this should be taken into account as a potential bias (Lincoln 2002).

Effects of interventions Incomplete outcome data In seven studies (35%), there were no losses to follow-up (Benedict 2000; Brissart 2012; Filippi 2012; Mattioli 2010; Mattioli 2012a;

See: Summary of findings for the main comparison Cognitive training versus any control for multiple sclerosis; Summary of findings 2 Cognitive training combined with other


15

neuropsychological rehabilitation methods versus any control for multiple sclerosis Eighteen of the 20 included studies showed some evidence of positive effects when the studies were individually analysed. In two studies (Lincoln 2002; Solari 2004) no significant effect of the intervention was reported.

sults. Everyday cognitive performance was not evaluated by these studies.

Immediate post-intervention assessment

Primary outcomes Cognitive training versus any control Nine original studies (Cerasa 2012; Filippi 2012; Mattioli 2010; Mattioli 2012a; Mendozzi 1998; Shatil 2010; Solari 2004; Vogt 2008; Vogt 2009) evaluated the effects of direct cognitive training. The target of the intervention was memory (Mendozzi 1998), working memory (Vogt 2008; Vogt 2009), memory and attention (Solari 2004), attention (Cerasa 2012), attention, information processing and executive functions (Filippi 2012; Mattioli 2010; Mattioli 2012a) and several cognitive functions based on the patient’s individual neuropsychological symptoms (Shatil 2010). Six of the studies (Cerasa 2012; Mattioli 2010; Mendozzi 1998; Shatil 2010; Solari 2004; Vogt 2008) were included in the metaanalyses. The studies Filippi 2012 and Mattioli 2012a were related to Mattioli 2010, and Vogt 2009 was related to Vogt 2008. These three studies (Filippi 2012; Mattioli 2012a; Vogt 2009) were not included in the meta-analyses to avoid the duplication of the re-

Analyses on the effects of cognitive training compared to any control (other intervention or no intervention) showed significant positive effects of cognitive training on memory span (standardised mean difference (SMD) fixed-effect 0.54, 95% confidence interval (CI) 0.20 to 0.88, P = 0.002, Analysis 1.3) and working memory (SMD fixed-effect 0.33, 95% CI 0.09 to 0.57, P = 0.006, Analysis 1.4). Cognitive training did not result in significant effects on attention (Analysis 1.1), information processing speed (Analysis 1.2), immediate verbal memory (Analysis 1.5), immediate visual memory (Analysis 1.6), delayed memory (Analysis 1.7), executive functions (Analysis 1.8) or verbal functions (Analysis 1.9). The corresponding funnel plot for attention (Figure 4) suggests possible publication bias, as there were two small positive studies but no small negative studies published (Higgins 2011). In other words, the two smaller, less precise studies were more positive than the larger, more precise studies.

Figure 4. Funnel plot of comparison: 1 Cognitive training versus any control, outcome: 1.1 Attention.


16

Secondary outcomes Analyses on the effects of cognitive training compared to any control showed no significant differences in depression (Analysis 1.10), quality of life (Analysis 1.11), fatigue (Analysis 1.12) or anxiety (Analysis 1.13).

Hildebrandt 2007; Lincoln 2002; Mäntynen 2013; Tesar 2005; Stuifbergen 2012) were included in the meta-analyses. The results of one study (Jonsson 1993) are reported in an additional table (Table 4), because data were not available in the appropriate format for meta-analyses.

Longitudinal assessment

Immediate post-intervention assessment

Neither of the above reported analyses showed significant differences at long-term follow-up (16 weeks). See Analysis 2.1; Analysis 2.2; Analysis 2.3; Analysis 2.4; Analysis 2.5; Analysis 2.6 and Analysis 2.7.

Primary outcomes

Cognitive training combined with other neuropsychological rehabilitation methods versus any control Nine original studies (Brissart 2012; Chiaravalloti 2005; Fink 2010; Hildebrandt 2007; Jonsson 1993; Lincoln 2002; Mäntynen 2013; Tesar 2005; Stuifbergen 2012) evaluated the effects of cognitive training combined with other neuropsychological methods. The interventions consisted of cognitive training and learning compensatory strategies. In these studies, the targets of the intervention were working memory, verbal learning, memory, attention, problem-solving skills and several cognitive functions. Eight studies (Brissart 2012; Chiaravalloti 2005; Fink 2010;

Analyses on the effects of cognitive training combined with other neuropsychological rehabilitation methods compared to any control (other intervention or no intervention) showed significant positive effects only for attention (SMD fixed-effect 0.15, 95% CI 0.01 to 0.28, P = 0.03, Analysis 3.1), immediate verbal memory (SMD fixed-effect 0.31, 95% CI 0.08 to 0.54, P = 0.008, Analysis 3.5) and delayed memory (SMD fixed-effect 0.22, 95% CI 0.02 to 0.42, P = 0.03, Analysis 3.7). The corresponding funnel plot for attention (Figure 5) did not suggest relevant publication bias. Cognitive training combined with other neuropsychological rehabilitation methods did not result in significant effects on information processing speed (Analysis 3.2), memory span (Analysis 3.3), working memory (Analysis 3.4), immediate visual memory (Analysis 3.6), executive functions (Analysis 3.8), visual functions (Analysis 3.9), or verbal functions (Analysis 3.10).


17

Figure 5. Funnel plot of comparison: 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control, outcome: 3.1 Attention.

Analyses on the effects of cognitive training combined with other neuropsychological rehabilitation methods compared to any control showed no significant differences for patient’s (Analysis 3.11) or carer’s (Analysis 3.12) reported everyday cognitive performance. Secondary outcomes

Analyses on the effects of cognitive training combined with other neuropsychological rehabilitation methods compared to any control showed no significant differences for depression (Analysis 3.13), quality of life (Analysis 3.14), fatigue (Analysis 3.15), anxiety (Analysis 3.16) or impact of the disease (Analysis 3.17). The other secondary outcome, personality disturbance, was not evaluated in these studies.

Longitudinal assessment

The same analyses at long-term follow-up (mean 25.3, range 11 to 52 weeks) showed significant positive effects of cognitive training combined with other neuropsychological rehabilitation methods on immediate visual memory (SMD fixed-effect 0.35, 95% CI 0.05 to 0.65, P = 0.02, Analysis 4.5) and delayed memory (SMD fixed-effect 0.32, 95% CI 0.10 to 0.55, P = 0.005, Analysis 4.6). Effects on other outcomes were not significant. See Analysis 4.1; Analysis 4.2; Analysis 4.3; Analysis 4.4; Analysis 4.7; Analysis 4.8; Analysis 4.9; Analysis 4.10; Analysis 4.11; Analysis 4.12; Analysis 4.13; Analysis 4.14; Analysis 4.15 and Analysis 4.16. The corresponding funnel plot for attention (Figure 6) suggests possible publication bias, as there was one small negative study, but no small positive studies published (Higgins 2011).


18

Figure 6. Funnel plot of comparison: 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up), outcome: 4.1 Attention.

Multimodal neuropsychological rehabilitation Two studies (Benedict 2000; Mendoza 2001) evaluated the effects of multimodal neuropsychological rehabilitation. In Benedict 2000, the target of the intervention was behavioural problems, while in Mendoza 2001 the use of a notebook was exploited in order to intensify the observation of the actual needs of institutionalised patients. In these studies interventions were not similar enough for the results to be pooled. Thus, we used narrative presentation to report the main results of these studies in an additional table (Table 5).

Sensitivity analyses Exclusion of studies in which concealment of allocation had been inadequate resulted in the exclusion of four studies (Mattioli 2010; Mendozzi 1998; Shatil 2010; Vogt 2008) and thus only two studies (Cerasa 2012; Solari 2004) were included in meta-analysis evaluating the effects of cognitive training. In studies evaluating the effects of cognitive training combined with other neuropsychological rehabilitation methods five studies (Brissart 2012;

Chiaravalloti 2005; Fink 2010; Hildebrandt 2007; Tesar 2005) were excluded and only three studies (Lincoln 2002; Mäntynen 2013; Stuifbergen 2012) were included in the meta-analysis. None of these analyses showed significant effects. Exclusion of studies in which outcome evaluation was not blinded or blinding remained unclear resulted in the exclusion of two studies (Shatil 2010; Vogt 2008) and the inclusion of four studies (Cerasa 2012; Mattioli 2010; Mendozzi 1998; Solari 2004) in the meta-analysis evaluating the effects of cognitive training. Analysis revealed significant positive effects of cognitive training on memory span (SMD fixed-effect 0.63, 95% CI 0.24 to 1.02, P = 0.001) and executive functions (MD fixed-effect 22.3, 95% CI 6.16 to 38.44, P = 0.007). In studies evaluating the effects of cognitive training combined with other neuropsychological rehabilitation methods one study (Tesar 2005) was excluded and seven studies (Brissart 2012; Chiaravalloti 2005; Fink 2010; Hildebrandt 2007; Lincoln 2002; Mäntynen 2013; Stuifbergen 2012) were included in the meta-analysis. Analysis revealed significant positive effects of cognitive training combined with other neuropsychological rehabilitation methods on attention (SMD fixed-effect 0.15, 95% CI 0.01 to 0.28, P = 0.03) and immediate verbal memory (SMD fixed-effect 0.31, 95% CI 0.07 to 0.55, P = 0.01).


19

Exclusion of studies in which loss to follow-up remained unclear or was greater than 20% resulted in the exclusion of one study (Shatil 2010) and the inclusion of five studies (Cerasa 2012; Mattioli 2010; Mendozzi 1998; Solari 2004; Vogt 2008) in the meta-analysis evaluating the effects of cognitive training. Analysis revealed significant positive effects of cognitive training on executive functions (MD fixed-effect 22.30, 95% CI 6.16 to 38.44, P = 0.007). In studies evaluating the effects of cognitive training combined with other neuropsychological rehabilitation methods two stud-

ies (Fink 2010; Hildebrandt 2007) were excluded and five studies (Brissart 2012; Chiaravalloti 2005; Lincoln 2002; Mäntynen 2013; Tesar 2005) were included in the meta-analysis. Analysis revealed significant positive effects of cognitive training combined with other neuropsychological rehabilitation methods on attention (SMD fixed-effect 0.15, 95% CI 0.01 to 0.29, P = 0.04) and immediate verbal memory (SMD fixed-effect 0.28, 95% CI 0.01 to 0.54, P = 0.04).


20


A D D I T I O N A L S U M M A R Y O F F I N D I N G S [Explanation]

Cognitive training combined with other neuropsychological rehabilitation methods versus any control for multiple sclerosis Patient or population: patients with multiple sclerosis Settings: outpatient Intervention: cognitive training combined with other neuropsychological rehabilitation methods versus any control Outcomes

Attention

Illustrative comparative risks* (95% CI)

Assumed risk

Corresponding risk

Control

Cognitive training combined with other neuropsychological rehabilitation methods versus any control

Relative effect (95% CI)

No of participants (studies)

The mean attention score The mean attention score SMD 0.15 (95% CI 0.01 894 in the control groups was in the intervention groups to 0.28), (5 studies) 8.7 was P = 0.03 0.15 standard deviations higher (0.01 to 0.28 higher)

Information processing The mean information speed processing score in the control groups was -608. 6

The mean information SMD 0.05 (95% CI -0.27 181 processing speed score to 0.36), (3 studies) in the intervention groups P = 0.77 was 0.05 standard deviations higher (0.27 lower to 0.36 higher)

Quality of the evidence (GRADE)

Comment

⊕⊕⊕ moderate1

⊕

very low2,3

Time-related scores were entered as negative variables.

21


Immediate verbal mem- The mean immediate ver- The mean immediate ver- SMD 0.31 (95% CI 0.08 308 ory bal memory score in the bal memory score in the to 0.54), (7 studies) control groups was 32.9 intervention groups was P = 0.008 0.31 standard deviations higher (0.08 to 0.54 higher)

⊕

very low2,3

Delayed memory

⊕⊕⊕ moderate1

The mean delayed mem- The mean delayed mem- SMD 0.22 (95% CI 0.02 400 ory score in the control ory score in the interven- to 0.42), (4 studies) groups was 9.5 tion groups was P = 0.03 0.22 standard deviations higher (0.02 to 0.42 higher)

Everyday cognitive per- The mean everyday cogformance/patient’s re- nitive performance/paport tient’s report score in the control groups was 6.0

22

The mean everyday cog- SMD 0.1 (95% CI -0.03 1011 nitive performance/pa- to 0.23), (4 studies) tient’s report score in the P = 0.13 intervention groups was 0.1 standard deviations higher (0.03 lower to 0.23 higher)

⊕⊕

low1,3,4

Depression

The mean depression The mean depression MD 0.25 (95% CI -2.11 to 187 score in the control score in the intervention 2.61), (4 studies) groups was -8.6 groups was P = 0.84 0.25 standard deviations higher (2.11 lower to 2.61 higher)

⊕⊕

low1,4

Depression scores were entered as negative variables.

Fatigue

The mean fatigue score in The mean fatigue score SMD 0.08 (95% CI -0.44 159 the control groups was - in the intervention groups to 0.59), (3 studies) 50.8 was P = 0.77 0.08 standard deviations higher (0.44 lower to 0.59

⊕

very low1,3,4

Fatigue scores were entered as negative variables.


higher) *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; MD: mean difference; SMD: standardised mean difference GRADE Working Group grades of evidence High quality: Further research is very unlikely to change our confidence in the estimate of effect. 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. 1 One

crucial limitation in the implementation of the study. Several crucial limitations in the implementation of the study. 3 Low number of participants. 4 High, unexplained heterogeneity. 2

23

DISCUSSION Summary of main results This review evaluated the effects of neuropsychological rehabilitation on MS based on randomised and quasi-randomised controlled trials. Twenty studies were identified in a comprehensive and systematic literature search and analysis, the sample consisting of a total of 966 MS participants. The review indicates that cognitive training improves memory span and working memory (see Summary of findings for the main comparison). However, although cognitive training showed significant effects on subcategories of cognitive performance, the majority of comparisons yielded no significant results. Cognitive training combined with other neuropsychological methods was found to improve attention, immediate verbal memory and delayed memory (see Summary of findings 2). In longitudinal follow-up, positive effects of cognitive training combined with other neuropsychological rehabilitation methods were found for immediate visual memory and delayed memory. There was no evidence of an effect of cognitive training or cognitive training combined with other neuropsychological rehabilitation methods on everyday cognitive functions, mood, fatigue, personality, anxiety or quality of life. However, although the pooled results in the meta-analyses yielded no significant findings, 18 out of the 20 studies showed positive effects when the results were individually evaluated. When interpreting the results, it must be taken into account that the cognitive training was mostly performed using computer-aided programs. The sensitivity analyses showed that exclusion of studies where concealment of allocation was not adequate resulted in the loss of the majority of data and consequent analyses. Exclusion of studies where outcome evaluation was not blinded or where data lost at follow-up was remarkable did not result in any significant changes to the results of the analyses.

Overall completeness and applicability of evidence Population, interventions and outcome measures in the studies

Population

The disease course, severity and duration, as well as the cognitive status of the MS patients in the included studies were heterogeneous. Over one-third of the studies (35%, n = 7) included MS patients despite their normal cognitive status. The primary aim of rehabilitation is to reduce existing symptoms or to support the management of them, so it is not meaningful to include patients without any cognitive problems. When including patients both

with and without cognitive impairments, evaluation of the effectiveness of rehabilitation can be distorted. However, most of the studies (65%, n = 13) included only MS participants with either self reported or documented cognitive impairment, or both. When relying only on self reports, it should be taken into account that MS patients’ self reports of their cognitive performance can be unreliable, a fact that has been noticed in many previous studies (e.g. Benedict 2003; Carone 2005; Sherman 2007). In the studies, MS participants were mainly handled as a single group without dividing them into cognitively more homogeneous subgroups. This may also distort and flatten the results, because the effects of the neuropsychological intervention are probably different in patients with different severities of cognitive impairment. Consequently, different kinds of rehabilitation procedures may be needed for them. Only in one study (and even then not until afterwards) were the results of the intervention inspected separately in patients with mild and moderate/severe cognitive impairment (Chiaravalloti 2005). The intervention was found to be especially effective when memory deficits were moderate to severe. Interventions

The interventions differed in all the studies included in the review. Moreover, the duration and frequency of the interventions varied significantly. In addition to heterogeneity between the studies, within-study variation also occurred. Jonsson (Jonsson 1993), Lincoln (Lincoln 2002) and Tesar (Tesar 2005) tailored the interventions according to each patient’s individual symptoms. Therefore, standardisation and direct comparison of the interventions was not possible. However, two subgroups of intervention types could be created: cognitive training and cognitive training combined with other neuropsychological rehabilitation methods, for example teaching compensatory strategies. This variability also reflects the everyday reality of clinical practice; due to the heterogeneity of MS patients’ neuropsychological symptoms, rehabilitation has to be tailored individually to each patient’s needs. However, research settings usually focus only on specific rehabilitation methods because these can be precisely described. These kinds of study settings are not necessarily able to take into account a patient’s individual needs and are not necessarily based on neuropsychological examination. Therefore, both the research settings and also the meta-analyses conducted may ignore the individual nature of neuropsychological rehabilitation and thus flatten the results. Additionally, in some studies (Brissart 2012; Benedict 2000; Cerasa 2012; Chiaravalloti 2005; Fink 2010; Jonsson 1993; Mendozzi 1998; Solari 2004) cognitive training was compared to other active treatments (e.g. specific versus unspecific cognitive training), thus masking potentially beneficial effects. When unspecific cognitive training is used as a control intervention, it should be taken into account that these tasks may partly promote the same functions as specific training. Although cognitive functions are theoretically distinguishable, they are highly complex, modular and interactive systems.


24

The number of intervention sessions varied from eight to 36, the duration of the rehabilitation intervention from four weeks to six months, and the frequency from two times per month to five times per week. When analysing the results with regard to the number of sessions, duration and frequency, no definite conclusions can be drawn about the effect of these factors on rehabilitation outcomes. Outcome measures

Careful selection of outcome measures is essential because elicitation of treatment effects is naturally only possible for the factors covered by the outcome measures. Effects not evaluated by the measurements remain undocumented even if they have occurred. Cognitive tests and various questionnaires were most often used as outcome measures in the included studies. Outcome measures were mainly commonly known, validated methods. However, both the cognitive tests and the questionnaires used varied considerably between studies. The most commonly used, in 50% of studies (n = 10), were the Paced Auditory Serial Addition Test (PASAT) and the Symbol Digit Modalities Test (SDMT). The Beck Depression Inventory (BDI) was used in 40% of studies (n = 8), three of them showing a positive response to rehabilitation. However, it should be taken into account that the BDI was not originally designed to be used to show the effects of rehabilitation but to evaluate the severity of depressive symptoms (Beck 1961). Because of the variety in the nature and aims of the neuropsychological interventions, it is only natural that a diversity of outcome measures was used. However, with multiple significance testing, the risk of making a type one error increases. In other words, there is a risk of incorrectly concluding that the intervention has had an effect. The use of a Bonferroni adjusted critical significance level can partly reduce this risk, but none of the studies adopted this method. Furthermore, the reliability of the findings can be lost if similar methods are used as training material and as outcome measures. For future meta-analyses, it would be beneficial if greater consensus could be achieved regarding the selection of standardised measures for use in outcome evaluation. Outcome measures should be focused on the functions targeted by rehabilitation so that the effects of the intervention programme can be realistically evaluated. Only three studies (Fink 2010; Mäntynen 2013; Solari 2004) separately specified the primary outcome measure and other secondary measures. Prior definition of one or two primary outcome measures could reduce the statistical problems of multiple comparisons and make the primary findings of the study more transparent. When evaluating the effects of a rehabilitation intervention, the aim should be to use outcomes which also measure, in addition to the severity of a symptom, its effects on everyday activities and how the patient copes with the symptom or problem. This is especially relevant for the progressive stage of disease. It is questionable whether changes in cognitive test performance also reflects real changes in everyday functions, when rehabilitation is based on individual aims relating to everyday functions (Wilson 2008). Even

in studies in which the main aim of the intervention is to teach compensatory strategies, the real benefits of the intervention do not necessarily become evident when using outcome measures that do not reflect the use of such strategies (Cicerone 2000). Ideally, outcome measures should reflect meaningful functional improvements such as the use of compensatory strategies to fulfil real-life demands, performance in everyday activities, changes in level of productivity or measures of subjective well-being (Cicerone 2000). The studies included in the present review were heterogeneous. Differences were obvious in the course and severity of the disease, as well as in the content, frequency, duration, setting and aim of the intervention, the outcome measures, the way of reporting the results and in the methodological quality of the studies. Because of this heterogeneity, it was not possible to pool the whole data set quantitatively with a meta-analysis. Meta-analysis was found to be appropriate in two subgroups: 1) studies comparing cognitive training versus any control and 2) studies comparing cognitive training combined with other neuropsychological rehabilitation methods versus any control.

Quality of the evidence The quality of the evidence for most of the outcomes was low or very low (see Summary of findings for the main comparison and Summary of findings 2). The strength of the findings of our review is limited due to the small sample sizes and methodological weaknesses of most of the studies. However, we did not exclude studies due to their low quality because the sample was already small. Instead, we evaluated the risk of bias of each study and the quality of the evidence. This enables us to report limited evidence based on lower-quality studies which, however, is the best evidence available at the present time. Most studies with two groups employed inappropriate randomisation methods. In these situations, the possible difference between the groups may be due to the biased selection of patients for the groups. Moreover, in several studies, the fact that the methods applied were not described in sufficient detail hampered our ’Risk of bias’ assessment. However, in the most recent studies it seems that quality has slightly improved and randomisation and allocation concealment have been more successful. The blinding of therapists or patients is usually not possible in research settings evaluating the effectiveness of rehabilitation due to the nature of the interventions. These deficiencies in blinding were also the most commonly occurring deficiencies in the evaluated research settings. However, blinding of the assessor or the person interpreting the results may be possible, and this can partly reduce the possibility of bias. This was slightly better realised in the studies, although only a single study reported how the blinding of the assessor was actually achieved. The quality domains best met in the included studies were incomplete outcome data and selective reporting, with only few studies failing to report appropriately.


25

Comprehensive ’Risk of bias’ (e.g. The Cochrane Collaboration’s ’Risk of bias’ assessment tool, Higgins 2011)) and quality assessment criteria (e.g. Van Tulder 2003) can act as a guide when constructing research settings. If the criteria for a high-quality study are taken into account in the planning of a study, the basis for a successful and effective study can be created. Fulfilling all the criteria is almost impossible with neuropsychological interventions, because the treatment is based on interaction between patients and healthcare professionals (e.g. the blinding of patient and therapist is rarely possible, unlike in studies of the effects of medications). Although it may be difficult to apply the existing quality criteria to studies on the effects of rehabilitation, it is nonetheless better to operate according to some consistent criteria than without any guidelines.

Potential biases in the review process We attempted to reduce bias in the review process by performing a comprehensive search and ensuring that two independent authors judged the relevance, risk of bias and content of the included studies. Although our search was comprehensive and we included studies identified in languages other than English, we cannot rule out the possibility that some studies have been missed. The possibility of publication bias may also have been increased by the fact that relevant studies with unclear or negative results may have remain unpublished. To be able to create a meaningful summary of numerous separate cognitive variables and to evaluate the effects of neuropsychological rehabilitation on cognitive domains, we classified the cognitive tests employed in the included studies into cognitive domains according to the functions they are traditionally used to measure (Lezak 2004). However, although the classification was based on a theoretical background, every classification is artificial and imperfect because cognitive tests are highly multifactorial and evaluate several cognitive functions. For the meta-analyses, we used test scores observed at the end of treatment. Thus, the results describe a clinical profile observed at a specific time point instead of a change observed after the intervention. It cannot be ruled out that possible baseline imbalances may have partly affected the results. Additionally, both authors of the review acted as authors in one of the evaluated studies (Mäntynen 2013).

Agreements and disagreements with other studies or reviews There are three previous systematic reviews evaluating the effects of neuropsychological rehabilitation in MS. In the Cochrane review by Thomas et al (Thomas 2006), the effects of psychological interventions in MS were evaluated based on the evidence from 16 randomised controlled trials (RCTs). However, only five of these trials were of neuropsychological interventions and therefore

the review did not comprehensively cover the literature on neuropsychological rehabilitation. This review concluded that there is some, although unclear, evidence for the effects of cognitive rehabilitation in MS. It was noted that the large number of outcome measures and small sample sizes hampered the interpretation. In the review of O’Brien et al (O’Brien 2008), the effects of cognitive rehabilitation in MS were evaluated based on the evidence from 16 trials (four RCTs, five controlled clinical trials (CCTs) and seven uncontrolled studies). According to the conclusions of this review, for MS patients’ verbal memory and learning, two rehabilitation methods can be recommended: 1) the story memory technique and 2) self generation (self generating/modifying the memorised material). According to this review, there is preliminary evidence, but it is not sufficient to make any recommendations about the effects of rehabilitation on attention and executive functions and general cognitive rehabilitation. In all, according to this review, cognitive rehabilitation in MS is in its infancy and more research is needed. In the Cochrane review of das Nair et al (das Nair 2012a), the effects of memory rehabilitation in MS were evaluated based on the evidence from eight RCTs. This review did not find evidence to support the effectiveness of memory rehabilitation on memory function or functional abilities in patients with MS. However, this conclusion was arrived at because of the limited quality of some of the primary studies. According to this review further, robust RCTs of higher methodological quality are needed. According to the present review there are preliminary positive findings for the effects of neuropsychological rehabilitation in MS. In line with the findings of the review of O’Brien et al (O’Brien 2008) our review indicates that cognitive training can in particular improve memory functions in addition to attention, while for other cognitive functions (information processing speed, executive, verbal and visual functions) no significant evidence of an effect was found. It should, however, be noted that none of the studies specifically aimed to improve those functions alone. Furthermore, consistent with previous reviews, the present review highlights the need for high-quality studies on the topic. The contradictory findings of our review and the review of das Nair et al (das Nair 2012a) may be related to the different inclusion criteria for studies as well as the fact that more original studies were included in our review.

AUTHORS’ CONCLUSIONS Implications for practice This review found low-level evidence for the positive effects of neuropsychological rehabilitation in MS. The interventions and the outcome measures included in the review were heterogeneous and consequently clinical inferences can only be drawn from single studies. New trials may therefore change the strength and direction


26

of the evidence. Decisions about neuropsychological rehabilitation for MS patients have to be based on theoretical and empirical knowledge in addition to research evidence.

Implications for research This review raises concerns about the quality of the included studies which evaluated the effects of neuropsychological rehabilitation in MS. In further studies it is important to include an objective baseline assessment of the cognitive status of patients and to evaluate the effects of the intervention in a sufficiently large group of cognitively homogeneous patients. Furthermore, the course and severity of the disease and other relevant disease variables should be clearly reported in the studies. This could offer important extra information about which MS participants benefit most from rehabilitation and to which patient populations the results may be generalised, thus helping to focus limited resources on those who need and could benefit from the interventions. The interventions should be reported in detail to enable more accurate evaluation and repetition. Additionally, a clear aim for the interventions should be defined beforehand. The primary outcome measure should assess whether the predefined aim of the intervention has been realised; in specific cognitive training this could be, for example, a few central cognitive variables targeted by the intervention. Alternatively, different questionnaires (mood, fatigue, meta-cognition etc.) can act as primary measures of more multimodal cognitive rehabilitation, depending on the targets of the intervention. One method could be to define the individual aim and to measure it using a scale such as Goal Attainment Scaling (GAS), for example. Furthermore, basic statistical information with exact numerical information and timing of assessment should also be reported for all outcome measures. Future studies should include measures for evaluating more extensive functional

capacity and, in this way, possibly uncover the generalised effects on the patient’s everyday functions. In addition to general measures, methods to enable us to measure the achievement of individual aims relating to everyday functions are needed. Future studies should also aim for longer follow-ups to determine the longerlasting effects of rehabilitation interventions. On the basis of the results of the current review, the actions specified in Table 6 are recommended to improve the quality of future studies. It should be kept in mind, however, that the currently widely used quality criteria (such as The Cochrane Collaboration’s ’Risk of bias’ tool (Higgins 2011) and those of Van Tulder et al (Van Tulder 2003)) are somewhat problematic in the evaluation of psychological interventions. For example, the blinding of patients and therapists is rarely possible. In future, the aim should be to modify the criteria to serve the research needs better and to raise the quality of psychological intervention studies. Regardless of the criteria used, unifying the assessment and reporting practices for establishing the effectiveness of interventions can be regarded as a gold standard. It is impossible to collect and analyse research evidence and conduct meta-analyses if the general recommendations for high-quality studies are not followed. More recent studies appear to have succeeded better than earlier studies in taking into account these quality criteria. One reason for this might be the fact that systematic reviews have made these criteria more visible and thus highlighted their importance.

ACKNOWLEDGEMENTS We thank the Cochrane Multiple Sclerosis and Rare Diseases of the Central Nervous System Group for their help and technical support in preparing the review.

REFERENCES

References to studies included in this review Benedict 2000 {published and unpublished data} Benedict RH, Shapiro A, Priore R, Miller C, Munschauer F, Jacobs L. Neuropsychological counseling improves social behavior in cognitively-impaired multiple sclerosis patients. Multiple Sclerosis 2000;6:391–6. Brissart 2012 {published data only} Brissart H, Leroy M, Morele E, Baumann C, Spitz E, Debouverie M. Cognitive rehabilitation in multiple sclerosis. Neurocase 2012 Aug 1 [Epub ahead of print]. Cerasa 2012 {published data only} Cerasa A, Gioia MC, Valentino P, Nistico R, Chiriaco C, Pirritano D, et al.Computer-assisted cognitive rehabilitation of attention deficits for multiple sclerosis: a randomized

trial with fMRI correlates. Neurorehabilitation and Neural Repair 2012;27(4):284–95. Chiaravalloti 2005 {published and unpublished data} Chiaravalloti ND, DeLuca J, Moore NB, Ricker JH. Treating learning impairments improves memory performance in multiple sclerosis: a randomised clinical trial. Multiple Sclerosis 2005;11:58–68. Filippi 2012 {published data only} Filippi M, Riccitelli G, Mattiloli F, Capra R, Stampatori C, Pagani E, et al.Multiple sclerosis: effects of cognitive rehabilitation on structural and functional MR imaging measures - an explorative study. Radiology 2012;262(3): 932–40. Fink 2010 {published data only} Fink F, Rischkau E, Butt M, Klein J, Eling P, Hildebrandt H. Efficacy of an executive function intervention programme


27

in MS: a placebo-controlled and pseudo-randomized study. Multiple Sclerosis 2010;16(9):1148–51.

double-blind controlled trial. Journal of the Neurological Sciences 2004;222:99–104.

Hildebrandt 2007 {published data only} Hildebrandt H, Lanz M, Hahn HK, Hoffmann E, Schwarze B, Schwendemann G, et al.Cognitive training in MS: effects and relation to brain atrophy. Restorative Neurology and Neuroscience 2007;25:33–43.

Stuifbergen 2012 {published data only} Stuifbergen AK, Becker H, Perez F, Morison J, Kullberg V, Todd A. A randomized controlled trial of a cognitive rehabilitation intervention for persons with multiple sclerosis. Clinical Rehabilitation 2012;26(10):882–93.

Jonsson 1993 {published data only (unpublished sought but not used)} Jonsson A, Korfitzen EM, Heltberg A, Ravnborg MH, Byskov-Ottosen E. Effects of neuropsychological treatment in patients with multiple sclerosis. Acta Neurologica Scandinavica 1993;88(6):394–400.

Tesar 2005 {published and unpublished data} Tesar N, Bandion K, Baumhackl U. Efficacy of a neuropsychological training programme for patients with multiple sclerosis - a randomised controlled trial. Wiener Klinische Wochenschrift 2005;117:747–54.

Lincoln 2002 {published and unpublished data} Lincoln NB, Dent A, Harding J, Weyman N, Nicholl C, Blumhardt LD, et al.Evaluation of cognitive assessment and cognitive intervention for people with multiple sclerosis. Journal of Neurology, Neurosurgery and Psychiatry 2002;72: 93–8.

Vogt 2008 {published data only} Vogt A, Kappos L, Stöcklin M, Gschwind L, Opwis K, Penner I-K. BrainStim - Wirksamkeit eines neu entwickelten kognitiven trainingsprogramms bei MS. Neurologie & Rehabilitation 2008;14(2):93–101.

Mäntynen 2013 {published data only} Mäntynen A, Rosti-Otajärvi E, Koivisto K, Lilja A, Huhtala H, Hämäläinen P. Neuropsychological rehabilitation does not improve cognitive performance but reduces perceived cognitive deficits in patients with multiple sclerosis: a randomized controlled multi-centre trial. Multiple Sclerosis Journal 2013 Jun 26 [Epub ahead of print]. Mattioli 2010 {published and unpublished data} Mattioli F, Stampatori C, Zanotti D, Parrinello G, Capra R. Efficacy and specificity of intensive cognitive rehabilitation of attention and executive functions in multiple sclerosis. Journal of the Neurological Sciences 2010;288:101–5. Mattioli 2012a {published data only} Mattioli F, Stampatori C, Scarpazza C, Parrinello G, Capra R. Persistence of the effects of attention and executive functions intensive rehabilitation in relapsing remitting multiple sclerosis. Multiple Sclerosis and Related Disorders 2012;1:168–73. Mendoza 2001 {published data only} Mendoza RJ, Pittenger DJ, Weinstein CS. Unit management of depression of patients with multiple sclerosis using cognitive remediation strategies: a preliminary study. Neurorehabilitation and Neural Repair 2001;15(1):9–14. Mendozzi 1998 {published data only} Mendozzi L, Pugnetti L, Motta A, Barbieri E, Gambini A, Cazzullo CL. Computer-assisted memory retraining of patients with multiple sclerosis. Italian Journal of Neurological Sciences 1998;19:431–8. Shatil 2010 {published and unpublished data} Shatil E, Metzer A, Horvitz O, Miller A. Home-based personalized cognitive training in MS patients: a study of adherence and cognitive performance. NeuroRehabilitation 2010;26:143–53. Solari 2004 {published data only} Solari A, Motta A, Mendozzi L, Pucci E, Forni M, Mancardi G, et al.Computer-aided retraining of memory and attention in people with multiple sclerosis: a randomised,

Vogt 2009 {published data only} Vogt A, Kappos L, Calabrese P, Stöcklin M, Gschwind L, Opwis K, et al.Working memory training in patients with multiple sclerosis - comparison of two different training schedules. Restorative Neurology and Neuroscience 2009;27: 225–35.

References to studies excluded from this review Allen 1995 {published data only} Allen DN, Longmore S, Goldstein G. Memory training and multiple sclerosis: a case study. International Journal of Rehabilitation and Health 1995;1(3):189–202. Allen 1998 {published data only} Allen DN, Goldstein G, Heyman RA, Rondinelli T. Teaching memory strategies to persons with multiple sclerosis. Journal of Rehabilitation Research and Development 1998;35(4):405–10. Altinkaya 2012 {published data only} Altinkaya A, Guclu I, Kurt E, Bingol A, Yandim Kuscu D, Sutlas PN, et al.Long-term effects of cognitive rehabilitation in multiple sclerosis: 12 months follow-up. European Journal of Neurology 2012;19:715. Basso 2006 {published data only} Basso MR, Lowery N, Ghormley C, Combs DM, Johnson J. Self-generated learning in people with multiple sclerosis. Journal of the International Neuropsychological Society 2006; 12:640–8. Basso 2007 {published data only} Basso MR, Ghormley C, Lowery N, Combs D, Bornstein RA. Self-generated learning in people with multiple sclerosis: an extension of Chiaravalloti and DeLuca (2002). Journal of Clinical and Experimental Neuropsychology 2007; 12:1–7. Ben 2012 {published data only} Ben AE, Hertzman G, Mosberg-Galili R, Hellmann M. Cognitive rehabilitation: functional strategies, problemsolving, and self-management skills for clients with multiple sclerosis. Multiple Sclerosis 2012;18:45–6.


28

Birnboim 2004 {published data only} Birnboim S, Miller A. Cognitive rehabilitation for multiple sclerosis patients with executive dysfunction. Journal of Cognitive Rehabilitation 2004;22:11–8. Bombardier 2008 {published data only} Bombardier CH, Cunniffe M, Wadhwani R, Gibbons LE, Blake KD, Kraft GH. The efficacy of telephone counseling for health promotion in people with multiple sclerosis: A randomised controlled trial. Archives of Physical Medicine and Rehabilitation 2008;89:1849–56. Brenk 2008 {published data only} Brenk A, Laun K, Haase CG. Short-term cognitive training improves mental efficiency and mood in patients with multiple sclerosis. European Neurology 2008;60:304–9. Brissart 2010 {published data only} Brissart H, Leroy M, Debouverie M. Cognitive rehabilitation in multiple sclerosis: preliminary results and presentation of a new program, PROCOG-SEP [Premiere evaluation d’un programme de remediation cognitive chez des patients atteints de sclerose en plaques: PROCOG–SEP]. Revue Neurologique 2010;166:406–11. Chiaravalloti 2002 {published data only} Chiaravalloti ND, DeLuca J. Self-generation as a means of maximizing learning in multiple sclerosis. Archives of Physical Medicine and Rehabilitation 2002;83(8):1070–9. Chiaravalloti 2003 {published data only} Chiaravalloti ND, Demaree H, Gaudino EA. Can the repetition effect maximize learning in multiple sclerosis?. Clinical Rehabilitation 2003;17(1):58–68. Chiaravalloti 2011 {published data only} Chiaravalloti ND. The modified story memory technique to treat learning and memory deficits in multiple sclerosis. Multiple Sclerosis 2011;17:43. Chiaravalloti 2012 {published data only} Chiaravalloti ND, Wylie G, Leavitt V, DeLuca J. Increased cerebral activation after behavioral treatment for memory deficits in MS. Journal of Neurology 2012;259:1337–46. das Nair 2012b {published data only} das Nair R, Lincoln NB. Evaluation of rehabilitation of memory in neurological disabilities (ReMiND): a randomized controlled trial. Clinical Rehabilitation 2012; 26(10):894–903. Foley 1994 {published data only} Foley FW, Dince WM, Bedell JR, LaRocca NG, Kalb R, Caruso LS, et al.Psychoremediation of communication skills for cognitively impaired persons with multiple sclerosis. Journal of Neurologic Rehabilitation 1994;8:165–76. Gordon 1997 {published data only} Gordon PA, Lam CS, Winter R. Interaction strain and persons with multiple sclerosis: effectiveness of a social skills program. Journal of Applied Rehabilitation Counselling 1997;28:5–11. Goverover 2008 {published data only} Goverover Y, Chiaravalloti N, DeLuca J. Self-generation to improve learning and memory of functional activities

in persons with multiple sclerosis: meal preparation and managing finances. Archives of Physical Medicine and Rehabilitation 2008;89(8):1514–21. Goverover 2011 {published data only} Goverover Y, Basso M, Wood H, Chiaravalloti N, DeLuca J. Examining the benefits of combining two learning strategies on recall of functional information in persons with multiple sclerosis. Multiple Sclerosis 2011;17(12):1488–97. Hubacher 2011 {published data only} Hubacher M, Kappos L, Opwis K, Stocklin M, Weier K, Sprenger T, et al.Benefits of cognitive training in MS patients treated with INFB-1b. Schweizer Archiv fur Neurologie und Psychiatrie 2011;162:6–7. Leavitt 2012 {published data only} Leavitt VM, Wylie GR, Girgis PA, DeLuca J, Chiaravalloti ND. Increased functional connectivity within memory networks following memory rehabilitation in multiple sclerosis. Brain Imaging and Behavior 2012 Jun 16 [Epub ahead of print]. Lincoln 2003 {published data only} Lincoln NB, Dent A, Harding J. Treatment of cognitive problems for people with multiple sclerosis. International Journal of Therapy and Rehabilitation 2003;10(9):412–5. Longley 2012 {published data only} Longley WA, Tate RL, Brown RF. A protocol for measuring the direct psychological benefit of neuropsychological assessment with feedback in multiple sclerosis. Brain Impairment 2012;13(2):238–55. Mattioli 2012b {published data only} Mattioli F, Stampatori C, Bellomi F, Capra R, Provinciali L, Compagnucci L, et al.Specific versus aspecific intensive cognitive training in MS: preliminary results of the SMICT study. Multiple Sclerosis 2012;18:48. Panicari 2012 {published data only} Panicari L, Rocca M, Valsasina P, Carla Riccitelli G, Mattioli F, Capra R, et al.Cognitive rehabilitation modifies functional connectivity of the anterior cingulate cortex in multiple sclerosis. Neurology 2012;78:78. Parisi 2012 {published data only} Parisi L, Rocca MA, Valsasina P, Panicari L, Mattioli F, Filippi M. Cognitive rehabilitation correlates with the functional connectivity of the anterior congulate cortex in patients with multiple sclerosis. Brain Imaging and Behavior 2012 Apr 19 [Epub ahead of print]. Pierfederici 2007 {published data only} Pierfederici L, Bacci L, Morici C Catena L, Viti B, Morgantini A. Efficacy of a verbal-auditory rehabilitative training for attentional deficits on attentional and mnestic ability in MS patients: a controlled-randomised clinical trial. Multiple Sclerosis 2007;13:230. Plohmann 1994 {published data only} Plohmann A, Kappos L, Brunnschweiler H. Evaluation of a computer-based attention retraining program for patients with multiple sclerosis. Schweizer Archiv für Neurologie und Psychiatrie 1994;145:35–6.


29


30

Plohmann 1998 {published data only} Plohmann AM, Kappos L, Ammann W, Thordai A, Wittwer A, Huber S, et al.Computer assisted retraining of attentional impairments in patients with multiple sclerosis. Journal of Neurology, Neurosurgery and Psychiatry 1998;64: 455–62.

Shevil 2013 {published data only} Shevil E. Development and evaluation of a cognitive rehabilitation program for persons with multiple sclerosis. ClinicalTrials.gov.

Rigby 2008 {published data only} Rigby SA, Thornton EW, Young CA. A randomised group intervention trial to enhance mood and self-efficacy in people with multiple sclerosis. British Journal of Health Psychology 2008;13:619–31.

Amato 2001 Amato MP, Ponziani G, Siracusa G, Sorbi S. Cognitive dysfunction in early-onset multiple sclerosis: a reappraisal after 10 years. Archives of Neurology 2001;58(10):1602–6.

Rodgers 1996 {published data only} Rodgers D, Khoo K, MacEachen M, Oven M, Beatty WW. Cognitive therapy for multiple sclerosis: a preliminary study. Alternative Therapies 1996;2:70–4. Sastre-Garriga 2010 {published data only} Sastre-Garriga J, Alonso J, Renom M, Arevalo MJ, Gonzalez I, Galan I, et al.A functional magnetic resonance proof of concept pilot trial of cognitive rehabilitation in multiple sclerosis. Multiple Sclerosis Journal 2010;17(4):457–67. Schwartz 1999 {published data only} Schwartz CE. Teaching coping skills enhances quality of life more than peer support: results of a randomised trial with multiple sclerosis patients. Health Psychology 1999;18: 211–20. Shevil 2009 {published data only} Shevil E, Finlayson M. Process evaluation of a selfmanagement cognitive program for persons with multiple sclerosis. Patient Education and Counseling 2009;76:77–83. Stuifbergen 2011 {published data only} Stuifbergen A, Becker H, Morgan S, Morrison J, Perez F. Home-based computer-assisted cognitive training: feasibility and perceptions of people with multiple sclerosis. International Journal of MS Care 2011;13:189–98. Tesar 2003 {published data only} Tesar N, Baumhackl U, Kopp M, Günther V. Effects of psychological group therapy in patients with multiple sclerosis. Acta Neurologica Scandinavica 2003;107:394–9. Topcular 2011 {published data only} Topcular B, Bingol A, Yildiz S, Tutuncu M, Demirci O, Saip S, et al.Long-term effects of cognitive rehabilitation in multiple sclerosis: six months follow-up. Multiple Sclerosis 2011;17:173–4. Wassem 2003 {published data only} Wassem R, Dudley W. Symptom management and adjustment of patients with multiple sclerosis: a 4-year longitudinal intervention study. Clinical Nursing Research 2003;12:102–17.

References to ongoing studies Brochet 2013 {published data only} Brochet B. Randomised controlled clinical trial of cognitive rehabilitation in multiple sclerosis and assessment by neuroimaging. ClinicalTrials.gov.

Additional references

Amato 2006 Amato MP, Zipoli V, Portaccio E. Multiple sclerosisrelated cognitive changes: a review of cross-sectional and longitudinal studies. Journal of the Neurological Sciences 2006;245(1-2):41–6. Amato 2008 Amato MP, Zipoli V, Portaccio E. Cognitive changes in multiple sclerosis. Expert Review of Neurotherapeutics 2008; 8(10):1585–96. Barten 2010 Barten LJ, Allington DR, Procacci KA, Rivey MP. New approaches in the management of multiple sclerosis. Drug Design, Development and Therapy 2010;24(4):343–66. Beck 1961 Beck AT, Ward CH, Mendelson M, Mock J, Erbaugh J. An inventory for measuring depression. Archives of General Psychiatry 1961;4:561–71. Benedict 2003 Benedict RH, Munschauer F, Linn R, Miller C, Murphy E, Foley F, et al.Screening for multiple sclerosis cognitive impairment using a self-administered 15-item questionnaire. Multiple Sclerosis 2003;9(1):95–101. Bonnet 2006 Bonnet MC, Deloire MS, Salort E, Dousset V, Petry KG, Brochet B. Evidence of cognitive compensation associated with educational level in early relapsing-remitting multiple sclerosis. Journal of the Neurological Sciences 2006;251(1-2): 23–8. Bowen 2007 Bowen A, Lincoln N. Cognitive rehabilitation for spatial neglect following stroke. Cochrane Database of Systematic Reviews 2007, Issue 2. [DOI: 10.1002/ 14651858.CD003586] Calabrese 2006 Calabrese P. Neuropsychology of multiple sclerosis - an overview. Journal of Neurology 2006;253(Suppl 1):110–5. Carone 2005 Carone DA, Benedict RH, Munschauer FE, Fishman I, Weinstock-Guttman B. Interpreting patient/informant discrepancies of reported cognitive symptoms in MS. Journal of the International Neuropsychological Society 2005; 11(5):574–83. Chiaravalloti 2008 Chiaravalloti ND, DeLuca J. Cognitive impairment in multiple sclerosis. Lancet Neurology 2008;7(12):1139–51.


31

Cicerone 2000 Cicerone KD, Dahlberg C, Kalmar K, Langenbahn DM, Malec JF, Bergquist TF, et al.Evidence-based cognitive rehabilitation: recommendations for clinical practice. Archives of Physical Medicine and Rehabilitation 2000;81: 1596–615. das Nair 2007 das Nair R, Lincoln N. Cognitive rehabilitation for memory deficits following stroke. Cochrane Database of Systematic Reviews 2007, Issue 3. [DOI: 10.1002/ 14651858.CD002293] das Nair 2012a das Nair R, Ferguson H, Stark DL, Lincoln NB. Memory Rehabilitation for people with multiple sclerosis. Cochrane Database of Systematic Reviews 2012, Issue 3. [DOI: 10.1002/14651858.CD008754] DeLuca 2004 DeLuca J, Chelune GJ, Tulsky DS, Lengenfelder J, Chiaravalloti ND. Is speed of processing or working memory the primary information processing deficit in multiple sclerosis?. Journal of Clinical and Experimental Neuropsychology 2004;26(4):550–62. DeLuca 2005 DeLuca J. Fatigue, cognition, and mental effort. In: J DeLuca editor(s). Fatigue as a Window to the Brain. Cambridge: Massachusetts Institute of Technology, 2005: 37–58. Feinstein 1992 Feinstein A, Kartsounis LD, Miller DH, Youl BD, Ron MA. Clinically isolated lesions of the type seen in multiple sclerosis: a cognitive, psychiatric, and MRI follow up study. Journal of Neurology, Neurosurgery and Psychiatry 1992;55 (10):869–76. Feinstein 2007 Feinstein A. The Clinical Neuropsychiatry of Multiple Sclerosis. New York: Cambridge University Press, 2007. Filippi 2010 Filippi M, Rocca MA. MRI and cognition in multiple sclerosis. Neurological Sciences 2010;31(2):231–4. Fischer 2001 Fischer, J. Cognitive impairment in multiple sclerosis. In: SD Cook editor(s). Handbook of Multiple Sclerosis. 3rd Edition. New York: Marcel Dekker, Inc, 2001:233–56. Gilewski 1990 Gilewski MJ, Zelinski EM, Schaie KW. The memory functioning questionnaire for assessment of memory complaints in adulthood and old age. Psychology and Aging 1990;5:482–90. GRADE 2008 Brozek J, Oxman A, Schünemann H. GRADEpro. 3.2 for Windows. Brozek J, Oxman A, Schünemann H, 2008. Gronwall 1977 Gronwall DM. Paced auditory serial-addition task: a measure of recovery from concussion. Perceptual and Motor Skills 1977;44:367–73.

Henry 2006 Henry JD, Beatty WW. Verbal fluency deficits in multiple sclerosis. Neuropsychologia 2006;44(7):1166–74. Higgins 2011 Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 [updated March 2011]. The Cochrane Collaboration, 2011. Available from www.cochrane-handbook.org. Hogan 1969 Hogan R. Development of an empathy scale. Journal of Consulting and Clinical Psychology 1969;33:307–16. Jennekens-Schinkel 1990 Jennekens-Schinkel A, Laboyrie PM, Lanser JB, van der Velde EA. Cognition in patients with multiple sclerosis after four years. Journal of the Neurological Sciences 1990;99(2-3): 229–47. Khan 2007 Khan F, Pallant J, Brand C. Caregiver strain and factors associated with caregiver self-efficacy and quality of life in a community cohort with multiple sclerosis. Disability and Rehabilitation 2007;30(16):1241–50. Kotterba 2003 Kotterba S, Orth M, Eren E, Fangerau T, Sindern E. Assessment of driving performance in patients with relapsing-remitting multiple sclerosis by a driving simulator. European Neurology 2003;50(3):160–4. Kujala 1997 Kujala P, Portin R, Ruutiainen J. The progress of cognitive decline in multiple sclerosis. A controlled 3-year follow-up. Brain 1997;120:289–97. Langdon 1999 Langdon DW, Thompson AJ. Multiple sclerosis: a preliminary study of selected variables affecting rehabilitation outcome. Multiple Sclerosis 1999;5(2): 94–100. Lezak 2004 Lezak MD, Howieson DB, Loring DW. Neuropsychological Assessment. New York: Oxford University Press, 2004. Lincoln 2000 Lincoln N, Majid M, Weyman N. Cognitive rehabilitation for attention deficits following stroke. Cochrane Database of Systematic Reviews 2000, Issue 4. [DOI: 10.1002/ 14651858.CD002842] Lincoln 2008 Lincoln NB, Radford KA. Cognitive abilities as predictors of safety to drive in people with multiple sclerosis. Multiple Sclerosis 2008;14(1):123–8. Mariani 1991 Mariani C, Farina E, Cappa SF, Anzola GP, Faglia L, Bevilacqua L, et al.Neuropsychological assessment in multiple sclerosis: a follow-up study with magnetic resonance imaging. Journal of Neurology 1991;238(7): 395–400. Mateer 2005 Mateer CA. Fundamentals of cognitive rehabilitation. In: Halligan PW, Wade DT editor(s). Effectiveness of


32

Rehabilitation for Cognitive Deficits. New York: Oxford University Press, 2005:21–30. McDonald 2001 McDonald WI, Compston A, Edan G, Goodkin D, Hartung HP, Lublin FD, et al.Recommended diagnostic criteria for multiple sclerosis: guidelines from the International Panel on the diagnosis of multiple sclerosis. Annals of Neurology 2001;50(1):121–7. O’Brien 2008 O’Brien AR, Chiaravalloti N, Goverover Y, Deluca J. Evidenced-based cognitive rehabilitation for persons with multiple sclerosis: a review of the literature. Archives of Physical Medicine and Rehabilitation 2008;89(4):761–9. Pantano 2006 Pantano P, Mainero C, Caramia F. Functional brain reorganization in multiple sclerosis: evidence from fMRI studies. Journal of Neuroimaging 2006;16(2):104–14. Penner 2005 Penner IK, Vogt A, Raselli C, Stoecklin M, Opwis K, Kappos L. The FSMC (Fatigue Scale for Motor and Cognitive Functions): a new patient-reported outcome measure for cognitive and motor fatigue in multiple sclerosis. Multiple Sclerosis 2005;11:264. Penner 2007 Penner I-K, Opwis K, Kappos L. Relation between functional brain imaging, cognitive impairment and cognitive rehabilitation in patients with multiple sclerosis. Journal of Neurology 2007;254(Suppl 2):1153–7. Polman 2005 Polman CH, Reingold SC, Edan G, Filippi M, Hartung HP, Kappos L, et al.Diagnostic criteria for multiple sclerosis: 2005 revisions to the “McDonald Criteria”. Annals of Neurology 2005;58(6):840–6. Poser 1983 Poser CM, Paty DW, Scheinberg L, McDonald WI, Davis FA, Ebers GC, et al.New diagnostic criteria for multiple sclerosis: guidelines for research protocols. Annals of Neurology 1983;13(3):227–31. RevMan 2012 Nordic Cochrane Centre, The Cochrane Collaboration. Review Manager (RevMan). 5.2. Copenhagen: Nordic Cochrane Centre, The Cochrane Collaboration, 2012. Rogers 2007 Rogers JM, Panegyres PK. Cognitive impairment in multiple sclerosis: evidence-based analysis and recommendations. Journal of Clinical Neuroscience 2007;14(10):919–27. Rosti 2007 Rosti E, Hämäläinen P, Koivisto K, Hokkanen L. Oneyear follow-up study of relapsing-remitting MS patients’ cognitive performances: Paced Auditory Serial Addition Test’s susceptibility to change. Journal of the International Neuropsychological Society 2007;13(5):791–8.

Sherman 2007 Sherman TE, Rapport LJ, Ryan KA. Awareness of deficit in multiple sclerosis. Journal of Clinical and Experimental Neuropsychology 2007;12:1–11. Smith 1973 Smith A. Symbol Digit Modalities Test: Manual. Los Angeles: Western Psychological Services, 1973. Sperling 2001 Sperling R, Guttmann C, Hohol M, Warfield S, Jakab M, Parente M, et al.Regional magnetic resonance imaging lesion burden and cognitive function in multiple sclerosis: a longitudinal study. Archives of Neurology 2001;58(1): 115–21. Spielberger 1983 Spielberger CD. Manual for the State-Trait Anxiety Inventory (STAI). Palo Alto, CA: Consulting Psychologists Press, 1983. Sumowski 2009 Sumowski JF, Chiaravalloti N, Deluca J. Cognitive reserve protects against cognitive dysfunction in multiple sclerosis. Journal of Clinical and Experimental Neuropsychology 2009; 31(8):913–26. Sunderland 1983 Sunderland A, Harris JE, Baddeley AD. Do laboratory tests predict everyday memory? A neuropsychological study. Journal of Verbal Learning and Behavior 1983;22(3):341–57. Thomas 2006 Thomas PW, Thomas S, Hillier C, Galvin K, Baker R. Psychological interventions for multiple sclerosis. Cochrane Database of Systematic Reviews 2006, Issue 1. [DOI: 10.1002/14651858.CD004431.pub2] Van Tulder 2003 Van Tulder M, Furlan A, Bombardier C, Bouter L. Updated method guidelines for systematic reviews in the Cochrane Collaboration Back Review Group. Spine 2003;28(12): 1290–9. Ware 1992 Ware JE, Sherbourne CD. The MOS 36 item short form health survey (SF-36). Medical Care 1992;30:349–63. Wechsler 1997a Wechsler D. Wechsler Adult Intelligence Scale. 3rd Edition. San Antonio, US: The Psychological Corporation, 1997. Wechsler 1997b Wechsler D. Wechsler Memory Scale. 3rd Edition. San Antonio, US: The Psychological Corporation, 1997. Wilson 1996 Wilson BA, Alderman N, Burgess PW, Emslie H, Evans JJ. Behavioural Assessment of the Dysexecutive Syndrome (BADS): Manual. Bury St Edmonds, England: Thames Valley Test Company, 1996. Wilson 2008 Wilson BA. Neuropsychological rehabilitation. Annual Review of Clinical Psychology 2008;4:141–62.


33

Zivadinov 2001 Zivadinov R, Sepcic J, Nasuelli D, De Masi R, Bragadin LM, Tommasi MA, et al.A longitudinal study of brain atrophy and cognitive disturbances in the early phase of relapsing-remitting multiple sclerosis. Journal of Neurology Neurosurgery and Psychiatry 2001;70(6):773–80.

References to other published versions of this review Rosti-Otajärvi 2011 Rosti-Otajärvi EM, Hämäläinen PI. Neuropsychological rehabilitation for multiple sclerosis. Cochrane Database of Systematic Reviews 2011, Issue 11. [DOI: 10.1002/ 14651858.CD009131] ∗ Indicates the major publication for the study


34

CHARACTERISTICS OF STUDIES

Characteristics of included studies [ordered by study ID] Benedict 2000 Methods

RCT; single-centre trial Country: USA Study dates not reported

Participants

N = 15; intervention: 8, control: 7. Additionally, healthy controls: 15 Age: intervention: 47.9 (6.6), control: 41.4 (12.2), healthy controls: 43.5 (9.4) Gender (female/male): intervention: 5/3, control: 5/2, healthy controls: 11/4 Education years: intervention: 14.3 (2.1), control: 13.4 (2.2), healthy controls: 13.9 (1. 8) MS disease course: intervention: 8 SP, control: 6 SP, 1 PP EDSS: intervention: 4.9 (2.2), control: 5.1 (2.6) Duration of disease (years): not reported Exclusion criteria: history of neurological disease other than MS. Drug or alcohol dependence. Psychiatric disease. Depressive disorder. Clinical MS relapse. Corticosteroid treatment within 3 weeks prior to participation

Interventions

Intervention: Neuropsychological Compensatory Training, NCT: aim - to enhance patient and caregiver understanding of MS-related deficits in cognitive functions, personality and behaviour, and to improve adaptive function by psychoeducational and cognitive-behavioural methods 3 main aims: 1) To teach the patient and caregiver to understand the neurological basis of cognitive impairment, pathological affect and abnormal social behaviour (in other words, abnormal behaviour is due to neurological injury) 2) To improve the patient’s capacity to appreciate the perspective of others (social skills training exercises, attentive listening, perspective taking, facilitative communication practiced by role-playing and audio-taping of spontaneous conversation) 3) To diminish the frequency of socially aggressive behaviour (cognitive-behavioural strategies taught to enhance patient self control and behaviour regulation) Number of sessions: 12 x 1 h. Duration: 12 weeks. Frequency: once per week Control: unspecific supportive psychotherapy, NSP: aim - to foster personality growth and behaviour change through emotional support and clinician empathy, and to help with stress modulation Number of sessions, duration and frequency: same as in intervention group

Outcomes

Primary and secondary outcomes: Beck Depression Inventory (BDI) Hogan Empathy Scale (HES) (caregiver reported) NEO-Personality Inventory (NEO-PI) Modified Social Aggression Scale (caregiver reported) Assessment timing: baseline and immediate follow-up (12 to 14 weeks)

Notes

Domain targeted: behaviour regulation


35

Benedict 2000

(Continued)

Risk of bias Bias

Authors’ judgement

Support for judgement

Random sequence generation (selection Low risk bias)

Assignment by random number table

Allocation concealment (selection bias)

Author’s definition: “Patients were assigned in sequence by a randomized schedule but not by a person independent of the eligibility determination and the allocation was not concealed”

High risk

Blinding (performance bias and detection Unclear risk bias) Participant

It remained unclear whether participants were blinded

Blinding (performance bias and detection High risk bias) Personnel

Personnel were not blinded

Blinding (performance bias and detection Low risk bias) Outcome assessor

Assessors were blinded

Incomplete outcome data (attrition bias) All outcomes

Low risk

Drop-outs: 0 (0%)

Selective reporting (reporting bias)

Low risk

The expected outcomes are reported

Other bias

Low risk

The study appears to be free of other sources of bias

Brissart 2012 Methods

Quasi-randomised trial; single-centre trial Country: France Study dates not reported

Participants

N = 20, intervention: 10, control: 10 Age: intervention: 42.5 (5.2), control: 41.3 (8.0) Gender: not reported Education years: intervention: 13.0 (2.2), control: 12.4 (1.7) MS disease course: 20 RR EDSS: intervention: 2.9 (1.4), control: 2.9 (1.2) Duration of disease (years): intervention: 5.0 (3.6), control: 7.2 (5.5) Exclusion criteria: no cognitive complaint or mild or moderate cognitive impairment. Not ambulatory. Not French native speaker. Other MS disease course than RRMS. EDSS


36

Brissart 2012

(Continued)

> 5.0. Disease duration > 20 years. Age over 50 years. Use of corticosteroids during the last 4 weeks before enrolment. Participation in another rehabilitation programme. Neuropsychological assessment done in the 3 months prior to inclusion. Other chronic or neurological disease than MS. Substance abuse Interventions

Intervention: ProCog-SEP program; aim - to train cognitive functions with a computer-aided program (ProCog-SEP) and to teach cognitive compensatory strategies In order to optimise meta-cognition, each cognitive function was explained at the beginning of each session, in order to link exercises with everyday life. Sessions included exercises about semantic memory and lexical access, visual and verbal episodic memory, working memory, associative memory, executive functions and several cognitive functions Number of sessions: 13 x 2 h. Duration: 6 months. Frequency: twice per month Control: discussion program; neutral discussions and various non-cognitive exercises Number of sessions, duration and frequency: same as in intervention group

Outcomes

Primary and secondary outcomes: Selective Reminding Test 10/36 Spatial Recall TEA/working memory, flexibility, incompatibility Semantic fluency Phonological fluency WAIS/digit span Boston Naming Test Assessment timing: baseline and immediate follow-up (6 months)

Notes

Cognitive domain targeted: several cognitive functions

Risk of bias Bias



Random sequence generation (selection High risk bias)

Patients were casually assigned by a blinded psychologist


Patients were casually assigned by a blinded psychologist

High risk

Blinding (performance bias and detection Low risk bias) Participant

Patients were not informed of the programme in which they were included






37

Brissart 2012

(Continued)


Low risk

Drop-outs: 0


Low risk


Other bias

Unclear risk

Gender distribution at baseline not reported

Cerasa 2012 Methods

RCT; single-centre trial Country: Italy Study dates not reported

Participants

N = 26, intervention: 13, control: 13 Age: intervention: 31.7 (9.2), control: 33.7 (10.3) Gender (female/male): intervention: 9/3, control: 8/3 Education years: intervention: 11, control: 12 MS disease course: all RR EDSS: intervention: 3, control: 2 Duration of disease (years): intervention: 4.3 (3.0), control: 5.1 (5.2) Exclusion criteria: severe cognitive impairment as evaluated by a detailed neuropsychological assessment. No predominant deficits in either attention and/or information processing speed, working memory and/or executive functions (failure in at least 1 of the following tests: SDMT, TM A-B, PASAT, Stroop word-colour task). Additional impairment in other cognitive domains. Relapse and steroid treatment for at least 1 month prior to study entry. Concomitant therapy with antidepressant or psychoactive drugs. EDSS > 4.0. History of psychiatric problems. Not optimal visual acuity

Interventions

Intervention: computer-aided attention training; aim - to train attentional functions with a computer-aided program (RehaCom modules divided attention, attention and concentration, and vigilance) Divided attention: the patient was required to stimulate a train driver and several distractions had to be taken into account, with increasing levels of difficulty Attention and concentration: an individual picture (target) was presented and then compared with a matrix of pictures and selected from the matrix, with increasing levels of difficulty Vigilance: the patient was trained to sustain his/her attention for a long period of time by providing response times limited to the various items. The task was to control a conveyor belt and to select the objects that differed from a sample in one or more details Number of sessions: 12 x 1 hour. Duration: 6 weeks. Frequency: twice per week Control: computer-aided visuomotor co-ordination tasks at home. Patients had to simply respond quickly and accurately to the appearance of target visual stimuli (numbers 2-4-6-8) on the screen by pressing the corresponding number key on the keyboard with increasing speed Number of sessions, duration and frequency: same as in intervention group


38

Cerasa 2012

(Continued)

Outcomes

Primary and secondary outcomes: Selective Reminding Test (SRT) 10/36 Spatial Recall Test (SPART) Symbol Digit Modalities Test (SDMT) Controlled Oral Word Association Test (COWAT) Paced Auditory Serial Addition Test (PASAT) Stroop Trail Making A+B (TM A-B) Beck Depression Inventory II (BDI-II) State-Trait Anxiety Inventory (STAI) Fatigue Severity Scale (FSS) MR imaging Assessment timing: baseline and immediate follow-up (6 weeks)

Notes

Cognitive domain targeted: attention

Risk of bias Bias




Assignment by computer-generated, sitestratified randomisation schedule


Participants were randomly allocated by assistants not involved in the study using a computer-generated allocation list

Low risk


Participants were blinded






Low risk

Drop-outs: 3 (12%) (intervention: 1, control: 2)


Low risk


Other bias

Low risk



39

Chiaravalloti 2005 Methods

Quasi-randomised trial; single-centre trial Country: USA Study was conducted between September 2000 and September 2001

Participants

N = 29, intervention: 15, control: 14 Age: intervention: 45.1 (13.8), control: 46.0 (9.3) Gender (female/male): intervention: 64% female, control: 57% female Education years: intervention: 14.6 (2.7), control: 15.0 (2.8) MS disease course: 17 RR, 4 PP, 7 SP Ambulation index: intervention: 3.2 (2.8), control: 2.4 (2.6) Duration of disease (years): intervention: 14.0 (8.4), control: 8.4 (5.0) Exclusion criteria: age over 69 years. History of neurological disorder other than MS. Alcohol or drug abuse. Bipolar disorder. Psychotic disorder. Schizophrenia. Head injury resulting in more than 30 minutes loss of consciousness. No MS relapse during 1 month. No problems in new learning (the performance of MS patients should be at least 1 standard deviation lower in BSR than the normative values). Impaired attention and verbal comprehension

Interventions

Intervention: aim - to strengthen the acquisition of new information into long-term memory by improving the quality of acquisition (context and imagery). This was aimed for through learning the Story Memory Technique (SMT) Patients were taught to: 1) Use visualisation (i.e. imagery) to facilitate new learning (sessions 1 to 4) 2) Utilise context to learn new information (e.g. story) even if information is seemingly unrelated (sessions 5 to 8) Number of sessions: 8 x 45 min. Duration: 4 weeks. Frequency: twice per week Control: same tasks as in intervention group but without Story Memory Technique (SMT) Reading a story and recalling as much as possible, after which they were asked specific questions about the story Number of sessions, duration and frequency: same as in intervention group

Outcomes

Primary and secondary outcomes: Hopkins Verbal Learning Test-revised (HVLT-R) Memory Functioning Questionnaire (MFQ) Beck Depression Inventory (BDI) State Trait Anxiety Inventory (STAI) Assessment timing: baseline, immediate follow-up (6 weeks) and longitudinal followup (11 weeks)

Notes

Cognitive domain targeted: verbal learning and learning

Risk of bias Bias




Support for judgement Assignment based on alternation (“odd and even numbers”)

40

Chiaravalloti 2005

(Continued)


High risk

Assignment based on alternation








Low risk



Low risk


Other bias

High risk

Significant difference between intervention and control groups in duration of disease

Filippi 2012 Methods

Same study as Mattioli 2010

Participants

See Mattioli 2010

Interventions

See Mattioli 2010. Aim of the present study - to evaluate brain changes after cognitive rehabilitation in patients with relapsing-remitting MS by using neuropsychologic assessment and structural and functional magnetic resonance (MR) imaging techniques

Outcomes

Same cognitive tests as in Mattioli 2010, additionally MR imaging

Notes

Cognitive function targeted: attention, information processing and executive functions

Risk of bias Bias







High risk


41

Filippi 2012

(Continued)

Blinding (performance bias and detection High risk bias) Participant

Participants were not blinded






Low risk

Drop-outs: 0 (0%)


Low risk


Other bias

Low risk


Fink 2010 Methods

Quasi-randomised trial; single-centre trial Country: Germany Study dates not reported

Participants

N = 50, intervention: 14, placebo: 17, control: 19 Age: 44.8 (8.2) Gender (female/male): 41/9 Education years: not reported MS disease course: 50 RR EDSS: not reported Duration of disease (years): 7.7 (7.7) Exclusion criteria: corticosteroid treatment during the last 4 weeks before enrolment. EDSS > 7. Neuropsychiatric disorder. Dementia. Relapse during the study phase

Interventions

Intervention: aim - to ease executive deficits by self training and receiving feedback and discussing the exercises with a psychologist Textbook exercises for executive functioning and meeting with a psychologist weekly for 1.5 hours to receive feedback and to discuss the exercises Number of sessions: 24 x 25 to 30 min. Duration: 6 weeks. Frequency: 4 times per week Placebo control: reaction capacity subtest of the computer-aided rehabilitation program (RehaCom). Participants had to respond fast and accurately to visual stimuli. They had to call the psychologist once a week to report the time having spent on training Number of sessions: 30 x 40 min. Duration: 6 weeks. Frequency: 5 times per week Control: no intervention


42

Fink 2010

(Continued)

Outcomes

Primary outcomes: Preference shifting, response shifting, 2-back (computer-based) Secondary outcomes: California Verbal Learning Test (CVLT) Wechsler Adult Intelligence Scale/short form (WIP) Expanded Disability Status Scale (EDSS) Multiple Sclerosis Functional Composite (MSFC) Brain parenchymal fraction (BRF) Assessment timing: baseline, immediate follow-up (6 weeks) and longitudinal followup (1 year)

Notes

Cognitive domain targeted: executive functions

Risk of bias Bias




Patients were appointed to 1 of 3 groups. In title mentioned “pseudo-randomised study”


Patients were appointed to 1 of 3 groups. In title mentioned “pseudo-randomised study”

High risk








High risk

Drop-outs: post-treatment: 10 (20%) (intervention: 3, placebo: 3, control: 4); 1year follow-up: 30 (60%) (intervention: 8, placebo: 9, control: 13)


High risk

Results for some of the outcome measures (short form of the Wechsler Adult Intelligence Scale and Multiple Sclerosis Functional Composite) not reported


43

Fink 2010

(Continued)

Other bias

Unclear risk

Exact baseline characteristics per groups were not reported; it remained unclear whether groups were similar e.g. with regards to education

Hildebrandt 2007 Methods

Quasi-randomised trial; single-centre trial Country: Germany Study dates not reported

Participants

N = 42, intervention: 17, control: 25 Age: intervention: 42.4, control: 36.5 Gender (female/male): intervention: 12/5, control: 13/12 Education years: intervention: 11.6, control: 11.2 MS disease course: intervention: 17 RR, control: 25 RR EDSS: intervention: 2.9, control: 2.7 Duration of disease (years): intervention: 5.4, control: 4.5 Exclusion criteria: EDSS over 7. Current or past medical illness or psychiatric disorder. Substance abuse. At least 4 weeks from corticosteroid treatment

Interventions

Intervention: aim - to strengthen working memory by direct exercises and teaching memory strategies. This was aimed for through computer-aided (VILAT-G 1.0) training Patients received a compact disk (CD) for home-based training to learn a word list. Subsequently, a series of calculation tasks were presented. The calculations relied heavily on working memory, because 3 numbers had to be added or subtracted without help of paper and pencil. Moreover, the result of the previous calculation had to be remembered and compared with the actual task. After series of calculations (2 to 15) the program asked the patient to recall the word list earlier learned. Typing the memorised words in a sequence that was structured by a common semantic category was rewarded by a special feedback sentence. The number of words and calculations were self adapted to the performance level Number of sessions: 30 x 30 min. Duration: 6 weeks. Frequency: 5 times per week The intensity or frequency of home-based training was not controlled Control: no intervention

Outcomes

Primary and secondary outcomes: Expanded Disability Status Scale (EDSS) Multiple Sclerosis Functional Composite (MSFC) including components: Nine-Hole Peg Test (9HPT), Timed 25-Foot Walk (TWT) and Paced Auditory Serial Addition Test (PASAT) California Verbal Learning Test (CVLT) Test battery of attention (TAP)/object alternation and alertness test Beck Depression Inventory (BDI) Fatigue Severity Scale (FSS) SF-36 Health Questionnaire (short form) MRI-parameters (brain atrophy) Assessment timing: baseline and immediate follow-up (8 weeks; 2 weeks post-treatment)


44

Hildebrandt 2007

(Continued)

Notes

Cognitive domain targeted: working memory, learning and memory recall

Risk of bias Bias







High risk



Blinding (performance bias and detection Low risk bias) Personnel

There were no personnel in this study. Training was carried out by participant him/herself with computer-aided program




Unclear risk

Drop-outs not reported


Low risk


Other bias

High risk

Significant difference between intervention and control groups in age and gender

Jonsson 1993 Methods

RCT; single-centre trial Country: Denmark Study dates not reported

Participants

N = 40, intervention: 20, control: 20 Age: intervention: 46.1 (7.3), control: 43.0 (9.0) Gender (female/male): intervention: 9/11, control: 10/10 Education years: intervention: 10.9 (2.0), control: 12.2 (2.9) MS disease course: 6 RR, 25 SP, 9 PP EDSS: intervention: 5.6 (1.7), control: 5.6 (1.8) Duration of disease (years): intervention: 15.0 (11.2), control: 15.1 (8.5) Exclusion criteria: neurological disease other than MS. Psychiatric disease unrelated to MS. Age over 60 years. Severe visual impairment. Severe motor dysfunction of arms or hands. Very severe cognitive impairment. History of alcohol or drug abuse. Regular


45

Jonsson 1993

(Continued)

use of psychopharmacia or analgesics. Foreigner. No cognitive impairment verified by neuropsychological testing Interventions

Intervention: aim - to alleviate the patient’s individual neuropsychological symptoms through cognitive training and neuropsychotherapy The cognitive training employed the common principles in cognitive rehabilitation programmes consisting of compensation, substitution and cognitive training Cognitive training of concentration was done with compiled, inverted and mirror-written text, “two-in-one” pictures, labyrinths etc. Memory was trained both directly and by learning compensatory strategies (visualisation using pictures of increasing complexity, shopping lists and appointments, calendar). The calendar was also used for planning daily activities. Stories were read aloud and the patient was encouraged to visualise and then retell the stories in order to learn how to structure a text Patients with visuospatial difficulties were trained partly with mosaic games, being corrected and urged to work slowly and systematically, and partly with practical exercises such as walking or wheelchair driving in and outside the hospital Along with the cognitive training, the patients took part in neuropsychotherapy to realise and accept their present cognitive and behavioural level of functioning, learning how to best use their available resources Number of sessions: 1 to 1.5 h each session, in all mean 17.3 h. Duration: about 6.5 weeks. Frequency: 3 times per week Control: unspecific mental stimulation: Watching and discussing different kinds of films. Reading and discussing newspaper articles. Playing games. Discussing personal problems and problems concerning disease acceptance Number of sessions: 1 to 1.5 h each session, in all mean 17.1 h. Duration and frequency: same as in intervention group

Outcomes

Primary and secondary outcomes: Wechsler Adult Intelligence Scale-revised (WAIS-R)/digit span Sentence repetition Associative learning List learning Visual gestalts, learning and retention. WAIS-R/Block Design Trail making Symbol Digit Modalities Test (SDMT) Street WAIS-R/Similarities WAIS-R/Picture arrangement Beck Depression Inventory (BDI) State-Trait Anxiety Inventory (STAI) Assessment timing: baseline, immediate follow-up (on day 45.6 (mean) after the 1st testing) and longitudinal follow-up (6 months)

Notes


Risk of bias Neuropsychological rehabilitation for multiple sclerosis (Review) Copyright © 2014 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd.

46

Jonsson 1993

(Continued)

Bias




Randomised assignment with sealed, opaque envelope system


Randomised assignment with sealed, opaque envelope system

Low risk

Blinding (performance bias and detection Unclear risk bias) Participant

It remained unclear whether participants were blinded



Blinding (performance bias and detection High risk bias) Outcome assessor

No patient was tested and treated by the same neuropsychologist. However, the authors realised quite early that the intended blind study was not possible, since just a few words from the patient indicated which treatment had been given


High risk



High risk

Pre-treatment evaluation was done with neuropsychological test battery including 41 measures of cognitive performance. However, post-treatment evaluation was done only with 25 cognitive tests. “We selected 25 cognitive tests from the original battery, excluding some tests with known high practice effect”

Other bias

Low risk


Lincoln 2002 Methods

RCT; single-centre trial Country: Great Britain Study dates not reported

Participants

N = 240, intervention: 79, assessment: 79, control: 82 Age (median): intervention: 43.0, assessment: 43.0, control: 40.5 Gender (female/male): intervention: 48/26, assessment: 56/16, control: 52/25 Education years: age left education (median): intervention: 16.0, assessment: 16.0,


47

Lincoln 2002

(Continued)

control: 16.0 MS disease course: 94 SP, 107 RR, 19 PP, 21 not known Ambulation Index (median): intervention: 4, assessment: 4, control: 3 Duration of disease (years): not reported Exclusion criteria: live over 20 mile radius from the hospital. Unable to co-operate with assessment at least 30 min at time. No consent to take part Interventions

Intervention: aim - to alleviate the patient’s individual neuropsychological symptom complex through cognitive training and using compensatory strategies Patients received detailed neuropsychological assessment. The information obtained was summarised for patients and, when the patients agreed, their relatives. Formal psychological reports were sent to the professionals involved in the patients’ care and to patients and their relatives Cognitive rehabilitation programme based on individually identified problems. This included learning both external (diaries, notebooks, calendars, lists) and internal (visualisation) memory strategies. More detailed information about intervention was not reported Number of sessions: not intensive, exact intensity or frequency, however, not reported. Duration: 4 months Assessment: same way as in intervention group, patients received detailed neuropsychological assessment lasting 3 hours with oral and written feedback Control: no intervention, 30 min screening assessment without feedback

Outcomes

Primary and secondary outcomes: General Health Questionnaire (GHQ) (also caregiver ratings) SF-36 and SF-54 Health Questionnaire (Overall Quality of Life (OQoL), Q53 and Satisfaction with Quality of Life (SQoL), Q54 Extended Activities of Daily Living Index (EADL) Everyday memory questionnaire (EMQ) (also caregiver ratings) Dysexecutive Syndrome Questionnaire (DEX) (also caregiver ratings) Memory Aids Questionnaire (MAQ) Guy’s Neurological Disability Scale (GNDS) Assessment timing: Baseline, immediate follow-up (4 months) and longitudinal followup (8 months)

Notes


Risk of bias Bias




Assignment with a computer-generated random number table


Participants were randomly allocated by an assistant not involved in the study using a computer-generated allocation list

Low risk


48

Lincoln 2002

(Continued)








Low risk

Drop-outs: 17 (7%) (intervention: 5, assessment: 7, control: 5)


High risk

The study did not include cognitive tests as outcome measures although intervention was cognitive rehabilitation and target cognitive training

Other bias

High risk

Compliance in the intervention was weak, only 46% of participants in the intervention group participated in the intervention as planned. This information became evident in later publication (2003) of authors

Mattioli 2010 Methods

Quasi-randomised trial; single-centre trial Country: Italy Participant examinations between June 2007 and December 2008

Participants

N = 20, intervention: 10, control: 10 Age (median): intervention: 42, control: 44 Gender (female/male): intervention: 10/0, control: 10/0 Education years (median): intervention: 8.0, control: 9.0 MS disease course: 20 RR EDSS (median): intervention: 2.5, control: 1.5 Duration of disease (years) (median): intervention: 16.5, control: 18.5 Exclusion criteria: 1 or more clinical exacerbations in the previous year. Loss of visual acuity. Ongoing major psychiatric disorder. Substance abuse. Mini Mental State Examination < 24. Intact performance in Paced Auditory Serial Addition Test 2“ and 3” and in Wisconsin Card Sorting Test

Interventions

Intervention: aim - to train attention, information processing and executive functions with a computer-aided program (RehaCom modules Plan a Day and Divided Attention) Module “Plan a day”: trains the patient’s ability to organise, plan and develop solution strategies employing realistic simulations of a set of scheduled dates and duties to be


49

Mattioli 2010

(Continued)

organised at specific places in a small city map Module “Divided attention”: the patient is required to simulate a train driver, carefully observing the control panel of the train and the countryside. Several distractions, such as crossing animals and train speed must be taken into account with increasing levels of difficulty Number of sessions: 36 x 1h. Duration: 12 weeks. Frequency: 3 times per week Control: no intervention Outcomes

Primary and secondary outcomes: Brief Repeatable Battery of Neuropsychological Tests (BRBNT) including: Selective Reminding Test (SRT), 10/36 Spatial Recall Test, Symbol Digit Modalities Test (SDMT), Paced Auditory Serial Addition Test 2’ and 3’ (PASAT), Controlled Oral Word Association Test (COWAT) Wisconsin Card Sorting Test (WCST) Test of Everyday Attention (TEA) Montgomery-Asberg Depression Rating Scale (MADRS) Multiple Sclerosis Quality of Life Questionnaire (MSQoL) Assessment timing: baseline, immediate follow-up (12 weeks)

Notes


Risk of bias Bias







High risk








Low risk

Drop-outs: 0 (0%)


Low risk


Other bias

Low risk



50

Mattioli 2012a Methods

Same study as Mattioli 2010 (longitudinal follow-up)

Participants

See Mattioli 2010

Interventions

See Mattioli 2010

Outcomes

See Mattioli 2010 Assessment timing: baseline, 3 months, 6 months

Notes


Risk of bias Bias







High risk








Low risk

Drop-outs: 0 (0%)


Low risk


Other bias

Low risk


Mendoza 2001 Methods

Quasi-randomised trial; single-centre trial Country: USA Study dates not reported

Participants

N = 20, intervention: 10, control: 10 Age: intervention: 54.6, control: 64.7 Gender (female/male): intervention: 7/3, control: 7/3


51

Mendoza 2001

(Continued)

Education years: not reported MS disease course: not reported EDSS: mentioned that groups did not differ, however data not reported Duration of disease (years): not reported. Had lived in the facility for mean 5 years (intervention) and 4.8 years (control) Exclusion criteria: primary admitting diagnosis not MS. Unable to read test stimuli. Diagnosed with a comorbid major mental disorder. Speed had deteriorated so that not able to answer test questions at sufficient verbal level. Performance on K-SNAP in the none impaired range. Unavailable at the time the protocol would be administered Interventions

Intervention: nursing staff who cared for the participants in the intervention group were educated in neuroanatomy related to cognitive dysfunction in MS, details of the project, rationale for using memory notebooks and participant’s neurocognitive strengths and weaknesses Each participant received a large notebook that was attached to the wheelchair. Certified nursing assistants were instructed to interview the participant to whom they had been assigned and determine whether the participant had any comments or concerns to be recorded. All this information was gathered during normal morning routine. The staff were encouraged to read and write down information in the notebooks Staff and researchers routinely reviewed the notebooks and marked the entries to indicate to the participant that his/her notebook was being read. When appropriate, the staff offered special assistance, as indicated by the participant’s notebook entries Number of sessions: neuropsychologist provided 4 x 1 h counselling for nursing staff. Intensity or frequency of other intervention not reported. Duration: 2 months Control: no intervention, normal treatment routines

Outcomes

Primary and secondary outcomes: Kaufman Short Neuropsychological Assessment (K-SNAP) Hopkins Verbal Learning Test (HVLT) North American New Adult Reading Test (NANART) Stroop Wechsler Memory Scale III (WMS-III)/digit span, family pictures, logical memory, letter-number sequencing Boston Naming Test Controlled Oral Word Association Test (COWAT) Beck Depression Inventory (BDI) Activity level (the number of unit events in which the participant took part) Assessment timing: baseline, immediate follow-up (2 months)

Notes


Risk of bias Bias




Support for judgement Randomly assigned to groups with the criterion of equal numbers of men and women in each condition

52

Mendoza 2001

(Continued)


High risk

Randomly assigned to groups with the criterion of equal numbers of men and women in each condition





Blinding (performance bias and detection Unclear risk bias) Outcome assessor

It remained unclear whether assessors were blinded


Low risk



High risk

Data for cognitive tests not reported

Other bias

High risk

Significant difference between intervention and control groups in age

Mendozzi 1998 Methods

Quasi-randomised trial; single-centre trial Country: Italy Participant recruitment between years 1994 and 1996

Participants

N = 60, intervention (specific training): 20, unspecific training: 20, control: 20 Age: intervention: 47.9 (9.4), unspecific training: 45.9 (12.1), control: 45.4 (6.8) Gender (female/male): intervention: 11/9, unspecific training: 12/8, control: 10/10 Education years: intervention: 12.7 (4.8), unspecific training: 13.0 (3.5), control: 11.7 (3.6) MS disease course: RR and SP (numbers not reported) EDSS: intervention: 3.7 (2.2), unspecific training: 4.0 (2.1), control: 3.3 (2.0) Duration of disease (years): intervention: 12.0 (7.7), unspecific training: 10.7 (7.6), control: 10.2 (6.9) Exclusion criteria: unstable clinical condition during 2 months prior to the first retraining session. Less than 5 years formal education. Insufficient visual function and manual dexterity to perform the neuropsychological tests. History or current clinical evidence of mental disorders. No subjective memory and attention deficits (verified also as objective memory, attention and reaction speed problems in test performances)

Interventions

Intervention: Specific Computer-assisted Memory Retraining Program, SCRP: aim - to train memory and attention with a computer-aided program (RehaCom) Memory: objects were displayed on the screen, the patient had to memorise the location


53

Mendozzi 1998

(Continued)

of objects. The objects were hidden. An object was displayed and the patient had to indicate the location of the hidden object corresponding to that displayed Attention: similar to memory task, except that objects were not hidden. The task was to locate the object exactly corresponding to the one displayed. At each response the patient received positive or negative feedback At the end of each session, patients were presented with a histogram summarising their performance. The psychologist briefly discussed the results with the patient and set the goals for the next session Number of sessions: 15 x 45 min. Duration: 8 weeks. Frequency: twice per week Unspecific training: unspecific Computer-assisted Retraining Program, NCRP. Visual tracking: the patient had to move a red dot into a blue circle by operating a joystick. As soon as the 2 overlapped, the blue circle started moving and the patient had to keep the red dot within the blue circle Reaction time: the patient had to press a key as quickly as possible on presentation of a target stimulus (go, no-go) At the end of each session patients were presented with a histogram summarising their performance. The psychologist briefly discussed the results with the patient and set the goals for the next session Number of sessions, duration and frequency: same as in intervention group Control: no intervention Outcomes

Primary and secondary outcomes: Wechsler Memory Scale (WMS)/logical memory, digit span, visual reproduction, verbal paired associates Corsi block’s test and recognition memory (computer-based) Memory Scale of the Luria-Nebraska Neuropsychological Battery (LNNB) Signal detection time (computer-based) Raven’s progressive matrices (PM) Assessment timing: baseline, immediate follow-up (average of 40 days post-treatment (14 weeks))

Notes

Cognitive domain targeted: memory

Risk of bias Bias




First 30 patients were randomly assigned to groups, while next 30 patients assigned to balance the groups as much as possible for sex, age and education


First 30 patients were randomly assigned to groups, while next 30 patients assigned to balance the groups as much as possible for sex, age and education

High risk


54

Mendozzi 1998

(Continued)



Blinding (performance bias and detection Unclear risk bias) Personnel

It remained unclear whether personnel were blinded




Low risk

Drop-outs: 1 (2%) (intervention: 1, unspecific training: 0, control: 0)


Low risk


Other bias

Low risk


Mäntynen 2013 Methods

RCT; multi-centre trial Country: Finland Participant recruitment between November 2010 and April 2011

Participants

N = 102, intervention: 60, control: 42 Age: intervention: 43.5 (8.7), control: 44.1 (8.8) Gender (female/male): intervention: 45/13, control: 31/9 Education years: intervention: 13.6 (2.3), control: 13.8 (2.6) MS disease course: 102 RRMS EDSS: intervention: 0 to 4: 93.1%; 4.5 to 5.5: 6.9%, control 0 to 4: 92.5%; 4.5 to 5. 5: 7.5% Duration of disease (years): intervention: 9.2 (6.6), control 10.1 (7.1) Exclusion criteria: other MS disease course than RRMS, EDSS > 6.0, no subjective attentional deficits (total score of questions 1, 2, 11 in the MSNQ-P < 6), not objective attentional deficits (SDMT < 50), age under 18 or over 59, history of alcohol or drug abuse, psychiatric disorder, acute relapses, neurological disease other than MS, ongoing neuropsychological rehabilitation, overall cognitive impairment (performance on all tests of the BRBNT under -1.5 SD compared to norms of healthy controls)

Interventions

Intervention: aim - to teach compensatory strategies by training attention and working memory with a computer-aided program (Foramen Rehab) and to offer psychoeducation and psychological support to promote coping with cognitive impairments The rehabilitation consisted of computer-based attention and working memory retraining, psychoeducation and teaching compensatory strategies, and homework assignments involving rehabilitation goals, as well as offering psychological support to promote coping with cognitive impairments. The computer-based training programme consisted of


55

Mäntynen 2013

(Continued)

ForamenRehab cognitive software attention and memory modules Number of sessions: 13 x 60 min. Duration: 13 weeks. Frequency: once per week Control: no intervention Outcomes

Primary outcomes: Perceived Deficits Questionnaire (PDQ) Symbol Digit Modalities Test (SDMT) Goal Attainment Scaling (GAS); only for the intervention group Secondary outcomes: Buschke Selective Reminding Test (BSR) 10/36 Spatial Recall Test Paced Auditory Serial Addition Test (2 and 3 seconds interstimulus) Controlled Oral Word Association Test (COWAT) Stroop Trail making A + B Beck Depression Inventory II (BDI-II) Multiple Sclerosis Impact Scale (MSIS-29) Multiple Sclerosis Neuropsychological Questionnaire - Patient (MSNQ-P) and Informant (MSNQ-I) Brief version of the World Health Organization Quality of Life (WHOQOL-Bref ) Fatigue Scale for Motor and Cognitive Fatigue (FSMC) Assessment timing: baseline, immediate follow-up (3 months) and longitudinal followup (6 months)

Notes

Cognitive domain targeted: attention and working memory

Risk of bias Bias




Assignment with a computer-generated random number table


Participants were randomly allocated by a statistician not involved in the study using a computer-generated allocation list

Low risk






Assessing neuropsychologists were blinded to the intervention, which was also shown in their estimates; only 61.7% of them being correct


56

Mäntynen 2013

(Continued)


Low risk

Drop-outs: 4 (4%) (intervention: 2, control: 2). The compliance with the intervention was 94.1%


Low risk

The study protocol was available

Other bias

Low risk


Shatil 2010 Methods

Quasi-randomised trial; single-centre trial Country: Israel The study was conducted between 14 November 2005 and 22 November 2006

Participants

N = 107, intervention: 59, control: 48 Baseline characteristics of study completers (N = 46, intervention: 22, control: 24) Age: intervention: 49.9 (1.9), control: 42.3 (10.7) Gender (female/male): intervention: 17/5, control: 19/5 Education: intervention: university 16, high school 6, control: university 17, high school 7 MS disease course: RR and RP (numbers not reported) EDSS: intervention: 2.6 (2.1), control: 2.5 (1.7) Duration of disease (years): not reported Exclusion criteria: other MS disease course than RR or RP. Healthy dominant hand not functioning. Did not speak Hebrew. Did not own or not able to use a personal computer at home. Did not express an interest in taking part in the study. Other neurological disease. Drug or alcohol abuse or dependence. Major depression. Known condition which required the use of psychotropic medication

Interventions

Intervention: aim - to explore unprompted adherence to a personalised, home-based, computer-aided cognitive training program (CogniFit Personal Coach, CPC) and to examine the impact of training on cognitive performance Patients received a CD containing the cognitive training program for home-based training. The selection of training tasks was determined by individual performance on the computer-based neuropsychological examination, hence no 2 people had the same training regimen. For each individual, the CPC assigned scores to 17 cognitive abilities that were subsequently trained by means of 21 different training tasks The degree of difficulty of the tasks was modified automatically according to the participant’s performance level Number of sessions: 24 x 20 to 30 min. Duration: 12 weeks. Frequency: 3 times per week Control: no intervention

Outcomes

Primary and secondary outcomes: Neuropsychological Examination - CogniFit Personal Coach (N-CPC) composing 15 tasks measuring a wide range of cognitive abilities such as memory, attention and eyehand co-ordination


57

Shatil 2010

(Continued)

Zung Depression Scale Expanded Disability Scale (EDSS) Fatigue Severity Scale (FSS) Assessment timing: baseline, immediate follow-up (12 weeks) Notes

Cognitive function targeted: several cognitive functions

Risk of bias Bias




A sufficient number of participants were first allocated for the intervention group. Participants with no home internet connection and remaining participants were allocated to the control group


A sufficient number of participants were first allocated for the intervention group. Participants with no home internet connection and remaining participants were allocated to the control group

High risk




There were no personnel in this study. Training was carried out by participant him/herself with computer-aided program




High risk

Drop-outs: 61 (57%) (intervention: 37, control: 24). Only 37% of participants completed entire training regimen


Low risk


Other bias

High risk



58

Solari 2004 Methods

RCT; single-centre trial Country: Italy Participant recruitment between January 2001 and May 2002

Participants

N = 82, intervention: 42, control: 40 Age: intervention: 46.2 (9.2), control: 41.2 (10.6) Gender (female/male): intervention: 26/14, control: 23/14 Education: intervention: elementary school 19, high school 21, control: elementary school 17, high school 20 MS disease course: 39 RR, 35 PR, 3 PP EDSS: intervention: 3.0, control: 4.0 Duration of disease (years since first symptoms): intervention: 15.2, control: 13.5 Exclusion criteria: age under 18 or over 65 years. MMSE < 24. Education less than 8 years. Ongoing major psychiatric disorder. 1 or more relapses in 3 months prior to enrolment. Immunomodulant or immunosuppressant treatment initiated in 4 months prior to enrolment. Cognitive rehabilitation in the 6 months prior to enrolment. No subjective experience of poor attention or memory and confirmed by a score below the 80th percentile in at least 2 components of the BRBNT

Interventions

Intervention: aim - to train memory and attention with a computer-aided program (RehaCom) Patients were treated on an individual basis as outpatients with computer-aided retraining program (RehaCom memory and attention retraining). The content of the intervention was not reported precisely Number of sessions: 16 x 45 min. Duration: 8 weeks. Frequency: twice per week Control: patients were treated on an individual basis as outpatients with computeraided retraining program (RehaCom visuo-constructional and visuomotor co-ordination retraining procedures). To minimise possible attentional and memory retraining, the visuomotor co-ordination procedure was simplified in that “puzzle” pieces were predisposed in the correct orientation and a black and white representation of the picture to be constructed was visible as an aid to puzzle completion Number of sessions, duration and frequency: same as in intervention group

Outcomes

Primary outcomes: Brief Repeatable Battery of Neuropsychological Tests (BRBNT) including: Paced Auditory Serial Addition Test (PASAT), Symbol Digit Modalities Test (SDMT), Selective Reminding, 10/36 Spatial Recall and Word List Generation Increase of 20% or more in at least 2 BRBNT test scores provided that there was no worsening by 20% or more in 2 or more of the remaining tests Secondary outcomes: Chicago Mood Depression Inventory (CMDI) 54-item MS Quality of Life Questionnaire (MSQOL-54) Assessment timing: baseline, immediate follow-up (8 weeks) and longitudinal followup (16 weeks)

Notes

Cognitive function targeted: attention and memory

Risk of bias


59

Solari 2004

(Continued)

Bias




Assignment with computer-generated random number table


Assignment with computer-generated random number table

Low risk



Blinding (performance bias and detection Unclear risk bias) Personnel

It remained unclear whether personnel were blinded




Low risk



Low risk


Other bias

High risk


Stuifbergen 2012 Methods

RCT; single-centre trial Country: USA Study dates not reported

Participants

N = 63, intervention: 36, control: 27 Baseline characteristics of study completers (N = 61, intervention: 34, control: 27) Age: all: 48.0 (8.8), no significant differences between intervention and control group Gender (female/male): intervention: 29/5, control: 25/2 Education: intervention: high school 13 (38%), associate school 3 (9%), bachelors school 8 (24%), graduate degree 10 (29%), control: high school 7 (26%), associate school 2 (7%), bachelors school 11 (41%), graduate degree 7 (26%) MS disease course: not reported EDSS: all: 5.2 (1.2) Duration of disease (years): all: 12.2 (7.4), no significant differences between intervention and control group Exclusion criteria: Age under 18 or over 60 years. Unable to understand and comply with the study protocol. Not clinically definite MS for at least 6 months. Unstable disease status at the time


60

Stuifbergen 2012

(Continued)

of the study entry. No subjective cognitive complaints confirmed by at least 5 ratings of “sometimes” or more often in the PDQ. Other medical causes of dementia. Other neurological disorder. Major psychiatric disorder. Major functional limitations that precluded them from participating in the study Interventions

Intervention: aim - to train cognitive functions with a computer-aided program as home training and to teach cognitive compensatory strategies in group sessions. The MAPSSMS (Memory, Attention, and Problem Solving Skills for Persons with Multiple Sclerosis) had 2 components: 1) weekly group sessions focused on the efficacy for use of cognitive compensatory strategies and 2) a computer-aided cognitive training program as home training. Group sessions provided information specific to MS-related cognitive difficulties and helped participants assess their difficulties and identify strategies to manage those difficulties and improve functional performance. The computer-aided program included internet-based game-like formats of attention (auditory and visual reactions to fixed and random points and divided attention), executive (organising information, attributes and groups, sequences, serial addition and simultaneous multiple addition), memory (auditory and visual sequenced and reversed sequenced recall, colour match, trail trace, sequenced blocks and recall for objects and locations) and problem-solving skills (puzzles requiring deductive reasoning and organising and analysing facts) Number of group sessions: 8 x 2 h. Duration: 8 weeks. Frequency: Once per week. Computer-aided cognitive home training: minimum of 45 minutes 3 times per week Control: no intervention

Outcomes

Primary and secondary outcomes: Minimal Assessment of Cognitive Function in MS (MACFIMS) including: Controlled Oral Word Association Test (COWAT), Judgement of Line Orientation test (JLO), California Verbal Learning Test (CVLT-II), Brief Visuospatial Memory Test - Revised (BVMT-R), Paced Auditory Serial Addition Test (PASAT), Symbol Digit Modalities Test (SDMT), Sorting Test from the Delis-Kaplan Executive Function System (DKEFS) Control Subscale of the Multiple Sclerosis Self Efficacy Scale (MSSE-Control) Strategy Subscale of the Multifactorial Memory Questionnaire (MMQ-Strategy) Multiple Sclerosis Neuropsychological Screening Questionnaire -patient (MSNQ-P)

Notes

Cognitive function targeted: memory, attention and problem solving skills

Risk of bias Bias




Assignment with computer-generated random number table and sealed envelopes


Participants were randomly allocated by a data analyst not involved in the study using a computer-generated random number table and sealed envelopes

Low risk


61

Stuifbergen 2012

(Continued)








Low risk



Low risk


Other bias

Low risk


Tesar 2005 Methods

Quasi-randomised trial; single-centre trial Country: Austria Study dates not reported

Participants

N = 20, intervention: 10, control: 10 Age: intervention: 45.3 (9.2), control: 46.9 (11.2) Gender (female/male): intervention: 7/3, control: 5/4 Education: not reported MS disease course: 13 RR, 6 SP EDSS: intervention: 4.5 (1.7), control: 4.4 (1.9) Duration of disease (years): intervention: 8.0 (4.2), control: 10.4 (7.1) Exclusion criteria: previous psychiatric history. Past drug or alcohol abuse. Other medical diagnosis than MS. Relapse during the past 30 days. Corticosteroid therapy. IQ below 85. Age over 60 years. Insufficient visual acuity. BDI > 11 scores. No mild or moderate cognitive impairments (cognitive impairment was not defined more precisely)

Interventions

Intervention: aim - to train the 2 cognitive areas most severely affected with a computeraided program (RehaCom) and to teach compensatory strategies relevant to everyday life. The content of the intervention concerning computer-aided training was not more specifically reported Compensation strategies: Memory and learning: internal and external memory supports, association mechanisms Attention: techniques for building up routines for behaviour and self control Problem-solving and planning: simplifying and visualisation Number of sessions: 12 x 1 h. Duration: 4 weeks. Frequency: 3 times per week Control: no intervention


62

Tesar 2005

(Continued)

Outcomes

Primary and secondary outcomes: Verbal learning test (VLT) Non-verbal learning test (NVLT) Sustained attention test (DAUF) Computer-aided card-sorting procedure (CKV) Mosaic test from the Hamburg Wechsler Intelligence Test-revised (HAWIE-R) Beck Depression Inventory (BDI) Modified Fatigue Impact Scale (MFIS) Follow-up questionnaire Assessment timing: baseline, immediate follow-up (4 weeks) and longitudinal followup (3 months)

Notes

Cognitive function targeted: several cognitive functions

Risk of bias Bias




Assignment to comparable groups with respect to clinical and demographic factors. Independent person determined which group acted as intervention and control group


Assignment to comparable groups with respect to clinical and demographic factors. Independent person determined which group acted as intervention and control group

High risk








Low risk



Low risk



63

Tesar 2005

Other bias

(Continued)

Low risk


Vogt 2008 Methods

Quasi-randomised trial; single-centre trial Country: Switzerland Study dates not reported

Participants

N = 50, intervention (MS): 15, control (MS): 15, intervention (healthy controls): 10, control (healthy controls): 10 Age: intervention (MS): 43.2 (8.8), control (MS): 46.3 (10.5), intervention (healthy controls): 39.5 (11.6), control (healthy controls): 43.4 (12.5) Gender (female/male): intervention (MS): 11/4, control (MS): 10/5, intervention (healthy controls): 7/3, control (healthy controls): 6/4 Education (0 = secondary school, 1 = college, 2 = university): intervention (MS): 1.6 (0. 5), control (MS): 1.5 (0.5), intervention (healthy controls): 1.7 (0.7), control (healthy controls): 1.6 (0.5) MS disease course: 22 RR, 7 SP, 1 PP EDSS: intervention (MS): 3.2 (1.8), control (MS): 3.2 (1.6) Duration of disease (years): intervention (MS): 9.1 (5.4), control (MS): 12.1 (9.0) Exclusion criteria: unstable symptomatic medication during previous months. Relapse or unstable disease phase within previous 3 months. Corticosteroid treatment within previous month. Other neurological or psychiatric disease. Healthy controls: neurological or psychiatric disease

Interventions

Intervention: aim - to train working memory with a computer-aided program (BrainStim) as home training Spatial orientation (module “City Map”): trains spatial orientation by either visual or verbal instructions to be remembered and finding the path using given arrows along virtual city map Visual object memory and the updating functions of working memory (module “Find Pairs”): the aim is to remember the location of cards that have been turned over and back again and find pairs of cards with the same image Verbal numeric working memory (module “Memorize Numbers”): numbers that are presented for a short time have to be remembered meanwhile performing an arithmetic distraction task The degree of difficulty of the tasks was modified automatically according to the participant’s performance level Intervention in question for both MS patients (n = 15) and healthy controls (n = 10) Number of sessions: 16 x 45 min. (each of the 3 modules 15 min). Duration: 4 weeks. Frequency: 4 times per week Control: no intervention

Outcomes

Primary outcomes: Wechsler Memory Scale (WMS-R)/digit span backwards, Corsi Block forwards Test battery of attention (TAP)/2-back Task (modified) Paced Auditory Serial Addition Test (PASAT) from Brief Repeatable Battery of Neu-


64

Vogt 2008

(Continued)

ropsychological Tests (BRB-N) Faces Symbol Test (FST) Symbol Digit Modalities Test (SDMT/from BRB-N) Fatigue Scale for Motor and Cognitive Fatigue (FSMC) Modified Fatigue Impact Scale (MFIS) Allgemeinen Depressionsskala (ADS) Functional Assessment of MS (FAMS) Secondary outcomes: Log files recorded during training (modules: City Map, Find Pairs and Memorise Numbers) Assessment timing: baseline (twice within 2 weeks), immediate follow-up (4 weeks) Notes

Cognitive function targeted: working memory

Risk of bias Bias




Assignment to comparable groups with regard to demographic factors (age, sex, education)



High risk




There were no personnel in this study. Training was carried out by the participant him/herself with a computer-aided program




Low risk

Drop-outs: 0 (0%)


Low risk


Other bias

Low risk



65

Vogt 2009 Methods

Same study as Vogt 2008

Participants

N = 45, intervention (high-intensity training): 15, intervention (distributed training): 15, control: 15 Age: intervention (high-intensity training): 43.2 (8.8), intervention (distributed training): 43.4 (12.3), control: 46.3 (10.5) Gender (female/male): intervention (high-intensity training): 11/4, intervention (distributed training): 9/6, control: 10/5 Education (0 = secondary school, 1 = college, 2 = university): intervention (high-intensity training): 1.6 (0.5), intervention (distributed training): 1.5 (0.5), control: 1.5 (0.5) MS disease course: 36 RR, 8 SP, 1 PP EDSS: intervention (high-intensity training): 3.2 (1.8), intervention (distributed training): 2.3 (1.1), control: 3.2 (1.6) Duration of disease (years): intervention (high-intensity training): 9.1 (5.4), intervention (distributed training): 8.1 (6.3), control: 12.1 (9.0) Exclusion criteria: relapse. Change in symptomatic medication within previous 3 months. Treatment with steroids within previous month. Other neurological or psychiatric disease

Interventions

Intervention: aim - to evaluate 2 different training schedules of a computer-aided (BrainStim) working memory training Same intervention methods as in the study of Vogt 2008. In this study computer-aided training either with high-intensity or with distributed training were compared Number of sessions: 16 x 45 min. (both training groups). Duration: 4 weeks (highintensity training) and 8 weeks (distributed training). Frequency: 4 times per week (highintensity training) and twice per week (distributed training) Control: no intervention

Outcomes

Same outcome methods as in the study of Vogt 2008 Assessment timing: baseline (twice within 2 weeks), immediate follow-up (for the highintensity training group and control group 4 weeks, for the distributed training group 8 weeks)

Notes

Cognitive domain targeted: working memory. This paper is based on the same study as Vogt 2008

Risk of bias Bias







High risk




66

Vogt 2009

(Continued)


There were no personnel in this study. Training was carried out by the participant him/herself with a computer-aided program




Low risk

Drop-outs: 0 (0%)


Low risk


Other bias

Low risk


BDI: Beck Depression Inventory BRBNT: Brief Repeatable Battery of Neuropsychological Test BSR: Buschke Selective Reminding Test EDSS: Expanded Disability Status Scale MMSE: Mini Mental State Examination MS: Multiple Sclerosis MSNQ-P: Multiple Sclerosis Neuropsychological Questionnaire - Patient MR: Magnetic Resonance (Imaging) PASAT: Paced Auditory Serial Addition Test PDQ: Perceived Deficits Questionnaire PP: Primary Progressive RR: Relapsing-Remitting RRSM: Relapsing-Remitting Multiple Sclerosis SD: Standard Deviation SDMT: Symbol Digit Modalities Test SP: Secondary progressive TM A-B: Trail Making A+B

Characteristics of excluded studies [ordered by study ID]

Study

Reason for exclusion

Allen 1995

Case report

Allen 1998

Not a controlled trial

Altinkaya 2012

Abstract

Basso 2006

Not a neuropsychological/cognitive intervention


67

(Continued)

Basso 2007


Ben 2012

Abstract

Birnboim 2004

Case report

Bombardier 2008


Brenk 2008

Controlled clinical trial (CCT)

Brissart 2010


Chiaravalloti 2002


Chiaravalloti 2003


Chiaravalloti 2011

Abstract

Chiaravalloti 2012

Results of neuropsychological outcomes not reported (same study as Leavitt 2012)

das Nair 2012b

Results of MS participants could not be separated from other neurological patient groups

Foley 1994

Case report

Gordon 1997


Goverover 2008


Goverover 2011

Within-group design

Hubacher 2011

Abstract

Leavitt 2012

Results of neuropsychological outcomes not reported (same study as Chiaravalloti 2012)

Lincoln 2003

Case report

Longley 2012

A protocol for an ongoing trial

Mattioli 2012b

Abstract

Panicari 2012

Abstract

Parisi 2012

Results of neuropsychological outcomes not reported (same study as Mattioli 2010 and Filippi 2012)

Pierfederici 2007

Abstract

Plohmann 1994

Preliminary report of a study in progress at the time of publication


68

(Continued)

Plohmann 1998


Rigby 2008


Rodgers 1996

Neuropsychological/cognitive rehabilitation could not be separated from the intervention

Sastre-Garriga 2010

Controlled clinical trial (CCT)

Schwartz 1999


Shevil 2009


Stuifbergen 2011

Qualitative, descriptive study

Tesar 2003


Topcular 2011

Abstract

Wassem 2003


MS: multiple sclerosis

Characteristics of ongoing studies [ordered by study ID] Brochet 2013 Trial name or title

Randomised controlled clinical trial of cognitive rehabilitation in multiple sclerosis and assessment by neuroimaging

Methods

RCT; single-centre trial Country: France

Participants

50 patients with MS (age 18 to 55, disease duration > 6 months and ≤ 15 years), 25 healthy controls

Interventions

Intervention: specific cognitive rehabilitation Rehabilitation will be focused on attention, executive functions, information processing speed and working memory based on the patient’s individual neuropsychological symptoms. Sessions will include varied exercised according to complexity and presentation modality, using computerised and “paper-and-pencil” tasks and meta-cognitive training Number of sessions: 50 x 1 h. Duration: 4 months. Frequency: 3 times per week Control: non-specific cognitive rehabilitation Group session without specific cognitive rehabilitation Number of sessions, duration, frequency: same as in intervention group Control: healthy participants No intervention. Same evaluation procedures as patients


69

Brochet 2013

(Continued)

Outcomes

Primary outcomes: the cognitive global executive z score (after 4 months) Secondary outcomes: brain activation, the z score of the SDMT, the daily-life cognitive questionnaire, clinical global impression of patients (after 4 and 8 months), the z cognitive global executive z score (after 8 months)

Starting date

May 2011

Contact information

Bruno Brochet: [email protected] Mathilde Deloire: [email protected]

Notes Shevil 2013 Trial name or title

Development and evaluation of a cognitive rehabilitation program for persons with multiple sclerosis

Methods

RCT; single-centre trial Country: Israel

Participants

50 patients with MS (18 years of age or older) with self reported cognitive difficulties

Interventions

Self management cognitive rehabilitation group intervention. 8-week cognitive rehabilitation programme facilitated by an occupational therapist. Programme goals include increased knowledge of cognitive impairments in MS, increased self efficacy to manage cognitive changes and increased use of cognitive compensatory strategies Control intervention: 8-week group programme that is not specifically directed to management of cognitive impairments

Outcomes

Primary outcomes: cognitive strategy use (pre-intervention, post-intervention, 3, 6, 12-month follow-up) Secondary outcomes: cognitive self efficacy (pre-intervention, post-intervention, 3, 6, 12-month follow-up)

Starting date

February 2011

Contact information

Eynat Shevil:[email protected]

Notes

MS: multiple sclerosis SDMT: Symbol Digit Modalities Test


70

DATA AND ANALYSES

Comparison 1. Cognitive training versus any control

Outcome or subgroup title 1 Attention 1.1 Paced Auditory Serial Addition Test 2’ 1.2 Paced Auditory Serial Addition Test 3’ 1.3 Symbol Digit Modalities Test 1.4 Divided attention 1.5 Avoiding distractions 1.6 Inhibition 1.7 Shifting attention 1.8 Trail Making A 1.9 Trail Making B 1.10 Stroop/colour-word 2 Information processing speed 2.1 Signal detection time/reaction time(s) 2.2 Response time 2.3 2-back (reaction time) 2.4 Test of Everyday Attention (median for auditory stimulus) 2.5 Test of Everyday Attention (median for visual stimulus) 3 Memory span 3.1 Corsi blocks forward 3.2 Digit span forward 4 Working memory 4.1 2-back (numbers correct) 4.2 Digit span backward 4.3 Corsi blocks backward 4.4 Verbal auditory working memory 4.5 Visual working memory 4.6 Auditory working memory 5 Immediate verbal memory 5.1 Buschke Selective Reminding Test (consistent long-term retrieval) 5.2 Wechsler Memory Scale/logical memory

No. of studies

No. of participants

5 2

573 97

Std. Mean Difference (IV, Fixed, 95% CI) Std. Mean Difference (IV, Fixed, 95% CI)

0.06 [-0.10, 0.23] 0.03 [-0.38, 0.44]

3

73

Std. Mean Difference (IV, Fixed, 95% CI)

0.35 [-0.12, 0.82]

4

150


0.02 [-0.30, 0.34]

1 1 1 1 1 1 1 4 1

46 46 46 46 23 23 23 176 60

Std. Mean Difference (IV, Fixed, 95% CI) Std. Mean Difference (IV, Fixed, 95% CI) Std. Mean Difference (IV, Fixed, 95% CI) Std. Mean Difference (IV, Fixed, 95% CI) Std. Mean Difference (IV, Fixed, 95% CI) Std. Mean Difference (IV, Fixed, 95% CI) Std. Mean Difference (IV, Fixed, 95% CI) Std. Mean Difference (IV, Random, 95% CI) Std. Mean Difference (IV, Random, 95% CI)

-0.05 [-0.63, 0.53] -0.05 [-0.63, 0.53] 0.23 [-0.35, 0.81] -0.14 [-0.72, 0.44] -0.28 [-1.10, 0.54] 0.01 [-0.81, 0.82] 0.54 [-0.29, 1.38] 0.15 [-0.33, 0.62] 0.80 [0.24, 1.36]

1 1 1

46 30 20

Std. Mean Difference (IV, Random, 95% CI) Std. Mean Difference (IV, Random, 95% CI) Std. Mean Difference (IV, Random, 95% CI)

0.17 [-0.41, 0.75] -0.02 [-0.74, 0.69] -0.77 [-1.68, 0.15]

1

20

Std. Mean Difference (IV, Random, 95% CI)

0.21 [-0.67, 1.09]

2 2 1 3 1 2 1 1

150 90 60 288 30 90 30 46

Std. Mean Difference (IV, Fixed, 95% CI) Std. Mean Difference (IV, Fixed, 95% CI) Std. Mean Difference (IV, Fixed, 95% CI) Std. Mean Difference (IV, Fixed, 95% CI) Std. Mean Difference (IV, Fixed, 95% CI) Std. Mean Difference (IV, Fixed, 95% CI) Std. Mean Difference (IV, Fixed, 95% CI) Std. Mean Difference (IV, Fixed, 95% CI)

0.54 [0.20, 0.88] 0.55 [0.11, 0.99] 0.52 [-0.02, 1.07] 0.33 [0.09, 0.57] -0.05 [-0.76, 0.67] 0.34 [-0.09, 0.78] 0.38 [-0.34, 1.10] 0.59 [0.00, 1.19]

1 1

46 46


0.52 [-0.07, 1.11] 0.08 [-0.50, 0.66]

4 3

360 120


0.20 [-0.02, 0.41] 0.26 [-0.10, 0.63]

1

60


-0.08 [-0.62, 0.46]

Statistical method


Effect size

71

5.3 Wechsler Memory Scale/verbal paired associations (easy) 5.4 Wechsler Memory Scale/verbal paired associations (hard) 5.5 Memory scale of the Luria-Nebraska Neuropsychological Battery 6 Immediate visual memory 6.1 Recognition memory (Corsi) 6.2 Wechsler Memory Scale/visual reproduction 6.3 10/36 Spatial Recall Test (immediate recall) 6.4 Faces Symbol Test 7 Delayed memory 7.1 Buschke Selective Reminding Test (delayed recall) 7.2 10/36 Spatial Recall Test (delayed recall) 8 Executive functions 8.1 Wisconsin Card Sorting Test (total errors) 8.2 Planning 8.3 Time estimation 9 Verbal functions 9.1 Naming 9.2 Controlled Oral Word Association Test (phonemic) 9.3 Controlled Oral Word Association Test (semantic) 10 Depression 10.1 Montgomery-Asberg Depression Rating Scale 10.2 Chicago Mood Depression Inventory 10.3 Allgemeine Depressionskala 10.4 Zung Depression Scale 10.5 Beck Depression Inventory II 11 Quality of life 11.1 Multiple Sclerosis Quality of Life 11.2 Functional Assessment of MS 11.3 54-item MS Quality of Life Questionnaire/mental health composite

1

60


0.29 [-0.25, 0.83]

1

60


0.31 [-0.23, 0.85]

1

60


0.11 [-0.43, 0.65]

5 1

270 60

Std. Mean Difference (IV, Random, 95% CI) Std. Mean Difference (IV, Random, 95% CI)

0.14 [-0.30, 0.58] 0.36 [-0.18, 0.91]

1

60


0.79 [0.23, 1.34]

3

120


-0.20 [-0.94, 0.54]

1 3 3

30 240 120

Std. Mean Difference (IV, Random, 95% CI) Std. Mean Difference (IV, Fixed, 95% CI) Std. Mean Difference (IV, Fixed, 95% CI)

0.06 [-0.66, 0.77] -0.19 [-0.45, 0.07] -0.07 [-0.43, 0.29]

3

120


-0.31 [-0.68, 0.05]

2 1

112 20


0.35 [-0.03, 0.73] 1.16 [0.20, 2.12]

1 1 4 1 2

46 46 186 46 97

Std. Mean Difference (IV, Fixed, 95% CI) Std. Mean Difference (IV, Fixed, 95% CI) Std. Mean Difference (IV, Fixed, 95% CI) Std. Mean Difference (IV, Fixed, 95% CI) Std. Mean Difference (IV, Fixed, 95% CI)

0.08 [-0.50, 0.66] 0.33 [-0.25, 0.91] 0.20 [-0.09, 0.49] 0.19 [-0.39, 0.77] 0.21 [-0.19, 0.61]

2

43


0.20 [-0.40, 0.80]

5 1

196 20


0.26 [-0.23, 0.75] 1.27 [0.29, 2.24]

1

77


0.0 [-0.45, 0.45]

1

30


0.05 [-0.66, 0.77]

1 1

46 23


-0.31 [-0.89, 0.27] 0.85 [-0.01, 1.71]

3 1

127 20


0.26 [-0.09, 0.61] 0.47 [-0.42, 1.36]

1

30


-0.13 [-0.84, 0.59]

1

77


0.36 [-0.09, 0.81]


72

12 Fatigue 12.1 Fatigue Scale for Motor and Cognitive functions 12.2 Modified Fatigue Impact Scale 12.3 Fatigue Severity Scale/raw score 12.4 Fatigue Severity Scale/delta score 13 Anxiety 13.1 State Trait Anxiety Inventory

3 1

129 30


0.13 [-0.22, 0.48] 0.18 [-0.54, 0.90]

1

30


0.19 [-0.53, 0.91]

1

46


-0.25 [-0.83, 0.33]

1

23


0.79 [-0.07, 1.64]

1 1

23 23

Mean Difference (IV, Fixed, 95% CI) Mean Difference (IV, Fixed, 95% CI)

4.40 [-3.87, 12.67] 4.40 [-3.87, 12.67]

Comparison 2. Cognitive training versus any control (longitudinal follow-up)

No. of studies

No. of participants

1 Attention 1.1 Paced Auditory Serial Addition Test 2’ 1.2 Symbol Digit Modalities Test 2 Immediate verbal memory

1 1

154 77


-0.22 [-0.54, 0.10] -0.38 [-0.83, 0.07]

1

77


-0.06 [-0.51, 0.39]

1

77

Mean Difference (IV, Fixed, 95% CI)

2.1 Buschke Selective Reminding Test/consistent long-term retrieval 3 Immediate visual memory 3.1 10/36 Spatial Recall Test/immediate recall 4 Delayed memory 4.1 Buschke Selective Reminding Test/delayed recall 4.2 10/36 Spatial Recall Test/delayed recall 5 Verbal functions 5.1 Controlled Oral Word Association Test (phonemic) 6 Depression 6.1 Chicago Mood Depression Inventory 7 Quality of life 7.1 54-item MS Quality of Life Questionnaire/mental health composite

1

77

Mean Difference (IV, Fixed, 95% CI)

-0.30 [-17.77, 17. 17] -0.30 [-17.77, 17. 17]

1 1

77 77


-1.20 [-4.86, 2.46] -1.20 [-4.86, 2.46]

1 1

154 77


-0.30 [-0.62, 0.02] -0.30 [-0.75, 0.15]

1

77


-0.30 [-0.75, 0.15]

1 1

77 77


2.10 [-1.87, 6.07] 2.10 [-1.87, 6.07]

1 1

77 77


-0.40 [-5.16, 4.36] -0.40 [-5.16, 4.36]

1 1

77 77


-9.40 [-28.38, 9.58] -9.40 [-28.38, 9.58]

Outcome or subgroup title

Statistical method


Effect size

73

Comparison 3. Cognitive training combined with other neuropsychological rehabilitation methods versus any control

Outcome or subgroup title 1 Attention 1.1 Sustained attention test/correct answers 1.2 Paced Auditory Serial Addition Test 3’/raw score 1.3 Paced Auditory Serial Addition Test 2’/raw score 1.4 Paced Auditory Serial Addition Test 3’/z-score 1.5 Test battery of attention/alertness without cueing 1.6 Symbol Digit Modalities Test 1.7 Trail Making A 1.8 Trail Making B 1.9 Stroop/colour-word interference time 1.10 Test of Everyday Attention/incompatibility 2 Information processing speed 2.1 Sustained Attention Test/variation reaction time 2.2 2-back/reaction time 2.3 Preference shifting/reaction time 2.4 Response shifting/reaction time 2.5 Test Battery of Attention/object alternation 3 Memory span 3.1 WAIS/digit span forward 4 Working memory 4.1 2-back/commissions 4.2 WAIS/digit span backward 4.3 Test of Everyday Attention/working memory 5 Immediate verbal memory 5.1 Verbal Learning Test 5.2 California Verbal Learning Test/learning trials 5.3 California Verbal Learning Test/total 5.4 Hopkins Verbal Learning Test - revised

No. of studies

No. of participants

5 1

894 19


0.15 [0.01, 0.28] -0.02 [-0.92, 0.88]

2

159


0.07 [-0.24, 0.39]

2

159


0.21 [-0.10, 0.53]

1

42


0.01 [-0.61, 0.62]

1

42


0.23 [-0.39, 0.85]

2

159


0.19 [-0.12, 0.51]

1 1 1

98 98 98

Std. Mean Difference (IV, Fixed, 95% CI) Std. Mean Difference (IV, Fixed, 95% CI) Std. Mean Difference (IV, Fixed, 95% CI)

0.26 [-0.14, 0.67] -0.04 [-0.44, 0.36] 0.11 [-0.29, 0.51]

1

20


0.63 [-0.27, 1.53]

3 1

181 19


0.05 [-0.27, 0.36] 0.0 [-0.90, 0.90]

1 1

40 40


0.25 [-0.44, 0.95] 0.07 [-0.62, 0.77]

1

40


0.25 [-0.45, 0.94]

1

42


-0.27 [-0.89, 0.35]

1 1 2 1 1 1

20 20 80 40 20 20

Mean Difference (IV, Fixed, 95% CI) Mean Difference (IV, Fixed, 95% CI) Std. Mean Difference (IV, Fixed, 95% CI) Std. Mean Difference (IV, Fixed, 95% CI) Std. Mean Difference (IV, Fixed, 95% CI) Std. Mean Difference (IV, Fixed, 95% CI)

0.40 [-0.46, 1.26] 0.40 [-0.46, 1.26] 0.25 [-0.22, 0.72] 0.24 [-0.46, 0.94] -0.20 [-1.08, 0.68] 0.75 [-0.16, 1.67]

7 1 2

308 19 82


0.31 [0.08, 0.54] 0.34 [-0.57, 1.25] 0.41 [-0.05, 0.88]

1

61


0.16 [-0.34, 0.67]

1

28


0.08 [-0.66, 0.82]

Statistical method


Effect size

74

5.5 Buschke Selective Reminding Test/consistent long-term retrieval 5.6 Buschke Selective Reminding Test/free recall 6 Immediate visual memory 6.1 Non-Verbal Learning Test 6.2 Brief Visuospatial Memory Test - revised/total 6.3 10/36 Spatial Recall/total correct 7 Delayed memory 7.1 California Verbal Learning Test/long delay free recall 7.2 Brief Visuospatial Memory Test - revised/delayed 7.3 Buschke Selective Reminding Test/delayed recall 7.4 10/36 Spatial Recall/delayed 8 Executive functions 8.1 Computer-aided Card-Sorting Procedure/correct answers 8.2 Preference shifting/trials to criterion 8.3 Response shifting/trials to criterion 8.4 Sorting Test from the Delis-Kaplan Executive Function System 8.5 Test of Everyday Attention/flexibility errors 9 Visual functions 9.1 Intelligence Test-revised (HAWIE-R) 9.2 Judgement of Line Orientation Test 10 Verbal functions 10.1 Controlled Oral Word Association Test (semantic fluency) 10.2 Controlled Oral Word Association Test (phonologic fluency) 10.3 Boston Naming Test 11 Everyday cognitive performance/patient’s report 11.1 Memory Aids Questionnaire

1

98


0.24 [-0.16, 0.65]

1

20


1.17 [0.21, 2.14]

4 1 1

198 19 61


0.36 [-0.14, 0.86] 0.34 [-0.57, 1.25] -0.04 [-0.54, 0.47]

2

118


0.77 [-0.49, 2.03]

4 2

400 103


0.22 [0.02, 0.42] 0.38 [-0.02, 0.77]

1

61


0.06 [-0.44, 0.57]

2

118


0.15 [-0.22, 0.52]

2

118


0.22 [-0.15, 0.59]

4 1

180 19


-0.08 [-0.39, 0.23] -0.02 [-0.92, 0.88]

1

40


-0.33 [-1.03, 0.37]

1

40


0.18 [-0.52, 0.87]

1

61


-0.30 [-0.81, 0.21]

1

20


0.57 [-0.32, 1.47]

2 1

80 19


0.40 [-0.04, 0.85] 0.79 [-0.16, 1.73]

1

61


0.29 [-0.21, 0.80]

3 3

219 179


0.26 [-0.01, 0.53] 0.23 [-0.07, 0.52]

1

20


0.05 [-0.83, 0.93]

1 4

20 1011


0.81 [-0.11, 1.73] 0.10 [-0.03, 0.23]

1

225


-0.08 [-0.36, 0.20]


75

11.2 Memory Functioning Questionnaire/overall rating 11.3 Everyday Memory Questionnaire/patient’s report 11.4 Dysexecutive Memory Questionnaire/patient’s report 11.5 Strategy Subscale of the Multifactorial Memory Questionnaire 11.6 Multiple Sclerosis Neuropsychological Questionnaire 11.7 Perceived Deficits Questionnaire 12 Everyday cognitive performance/carer’s report 12.1 Everyday Memory Questionnaire/carer’s report 12.2 Dysexecutive Syndrome Questionnaire/carer’s report 12.3 Multiple Sclerosis Neuropsychological Questionnaire - Informant 13 Depression 13.1 Beck Depression Inventory 14 Quality of life 14.1 SF-36 Health Questionnaire (short form)/mental score 14.2 Brief version of the World Health Organization Quality of Life/psychological total score 15 Fatigue 15.1 Modified Fatigue Impact Scale 15.2 Fatigue Severity Scale 15.3 Fatigue Scale for Motor and Cognitive Fatigue/total 16 Anxiety 16.1 State-Trait Anxiety Inventory/state 17 Impact of the disease 17.1 Multiple Sclerosis Impact Scale/psychological total score 17.2 Control subscale of the Multiple Sclerosis Self Efficacy Scale

1

28


-0.40 [-1.15, 0.35]

1

217


-0.02 [-0.31, 0.26]

1

223


0.01 [-0.27, 0.30]

1

61


0.30 [-0.21, 0.81]

2

159


0.26 [-0.06, 0.58]

1

98


0.69 [0.27, 1.10]

2

454


0.03 [-0.17, 0.23]

1

177


0.04 [-0.28, 0.36]

1

179


-0.01 [-0.33, 0.31]

1

98


0.08 [-0.32, 0.49]

4 4

187 187


0.25 [-2.11, 2.61] 0.25 [-2.11, 2.61]

2 1

140 42


0.16 [-0.18, 0.50] 0.06 [-0.56, 0.68]

1

98


0.20 [-0.20, 0.61]

3 1

159 19


0.08 [-0.44, 0.59] -0.51 [-1.43, 0.41]

1 1

42 98


-0.05 [-0.66, 0.57] 0.43 [0.03, 0.84]

1 1

28 28


2.36 [-4.81, 9.53] 2.36 [-4.81, 9.53]

2 1

159 98


0.04 [-0.27, 0.36] 0.03 [-0.38, 0.43]

1

61


0.07 [-0.44, 0.58]


76

Comparison 4. Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up)

Outcome or subgroup title 1 Attention 1.1 Sustained attention test/correct answers 1.2 Paced Auditory Serial Addition Test 3’ 1.3 Paced Auditory Serial Addition Test 2’ 1.4 Symbol Digit Modalities Test 1.5 Trail Making A 1.6 Trail Making B 1.7 Stroop/colour-word interference time 2 Information processing speed 2.1 Sustained attention test/variation reaction time 2.2 2-back/reaction time 2.3 Preference shifting/reaction time 2.4 Response shifting/reaction time 3 Working memory 3.1 2-back/commissions 4 Immediate verbal memory 4.1 Verbal Learning Test 4.2 Hopkins Verbal Learning Test - revised 4.3 California Verbal Learning Test/learning trials 4.4 California Verbal Learning Test/total 4.5 Buschke Selective Reminding Test/consistent long-term retrieval 5 Immediate visual memory 5.1 Non-Verbal Learning Test 5.2 Brief Visuospatial Memory Test - revised/total 5.3 10/36 Spatial Recall/total correct 6 Delayed memory 6.1 California Verbal Learning Test/delay

No. of studies

No. of participants

3 1

790 19


0.01 [-0.13, 0.15] -0.65 [-1.58, 0.28]

2

159


0.18 [-0.14, 0.49]

2

159


-0.04 [-0.36, 0.28]

2

159


0.11 [-0.21, 0.42]

1 1 1

98 98 98


-0.10 [-0.50, 0.31] -0.10 [-0.50, 0.30] -0.01 [-0.41, 0.40]

2 1

79 19


0.05 [-0.43, 0.53] 0.56 [-0.36, 1.49]

1 1

20 20


-0.51 [-1.49, 0.46] -0.04 [1.00, 0.91]

1

20


0.14 [-0.82, 1.09]

1 1 5 1 1

20 20 226 19 28

Mean Difference (IV, Fixed, 95% CI) Mean Difference (IV, Fixed, 95% CI) Std. Mean Difference (IV, Fixed, 95% CI) Std. Mean Difference (IV, Fixed, 95% CI) Std. Mean Difference (IV, Fixed, 95% CI)

1.1 [-3.22, 5.42] 1.1 [-3.22, 5.42] 0.26 [-0.01, 0.53] 0.47 [-0.45, 1.38] -0.12 [-0.87, 0.62]

1

20


0.41 [-0.55, 1.38]

1

61


0.33 [-0.18, 0.84]

1

98


0.26 [-0.14, 0.67]

3 1 1

178 19 61


0.35 [0.05, 0.65] 0.05 [-0.85, 0.95] 0.05 [-0.46, 0.55]

1

98


0.62 [0.21, 1.03]

2 1

318 61


0.32 [0.10, 0.55] 0.26 [-0.24, 0.77]

Statistical method


Effect size

77

6.2 Brief Visuospatial Memory Test - revised/delayed 6.3 Buschke Selective Reminding Test/delayed 6.4 10/36 Spatial Recall/delayed 7 Executive functions 7.1 Computer-aided Card-Sorting Procedure/correct answers 7.2 Preference shifting/trials to criterion 7.3 Response shifting/trials to criterion 7.4 Sorting Test from the Delis-Kaplan Executive Function System 8 Visual functions 8.1 Intelligence Test-revised (HAWIE-R) 8.2 Judgement of Line Orientation Test 9 Verbal functions 9.1 Controlled Oral Word Association Test (semantic fluency) 10 Everyday cognitive performance/patient’s report 10.1 Memory Aids Questionnaire 10.2 Memory Functioning Questionnaire/overall memory 10.3 Everyday Memory Questionnaire/patient’s report 10.4 Dysexecutive Syndrome Questionnaire/patient’s report 10.5 Strategy Subscale of the Multifactorial Memory Questionnaire 10.6 Multiple Sclerosis Neuropsychological Questionnaire 10.7 Perceived Deficits Questionnaire 11 Everyday cognitive performance/carer’s report 11.1 Everyday Memory Questionnaire/carer’s report 11.2 Dysexecutive Syndrome Questionnaire/carer’s report

1

61


0.20 [-0.30, 0.71]

1

98


0.20 [-0.21, 0.60]

1

98


0.57 [0.16, 0.99]

3 1

120 19


-0.18 [-0.69, 0.32] -0.65 [-1.58, 0.28]

1

20


0.75 [-0.24, 1.74]

1

20


-0.37 [-1.34, 0.59]

1

61


-0.29 [-0.80, 0.22]

2 1

80 19


0.09 [-0.35, 0.53] 0.07 [-0.83, 0.98]

1

61


0.10 [-0.41, 0.60]

2 2

159 159


0.95 [-1.75, 3.65] 0.95 [-1.75, 3.65]

4

978


0.11 [-0.12, 0.33]

1

209


-0.09 [-0.38, 0.20]

1

28


-0.43 [-1.18, 0.32]

1

208


0.01 [-0.28, 0.31]

1

215


-0.01 [-0.30, 0.27]

1

61


0.17 [-0.34, 0.68]

2

159


0.16 [-0.51, 0.83]

1

98


0.73 [0.32, 1.15]

2

427


0.03 [-0.17, 0.23]

1

163


0.04 [-0.29, 0.36]

1

166


-0.03 [-0.36, 0.29]


78

11.3 Multiple Sclerosis Neuropsychological Questionnaire - Informant 12 Depression 12.1 Beck Depression Inventory 13 Quality of life 13.1 Brief version of the World Health Organization Quality of Life/psychological total score 14 Fatigue 14.1 Fatigue Impact Scale 14.2 Fatigue Scale for Motor and Cognitive Fatigue/total score 15 Anxiety 15.1 State Trait Anxiety Inventory/state 16 Impact of the disease 16.1 Multiple Sclerosis Impact Scale/psychological total score 16.2 Control subscale of the Multiple Sclerosis Self Efficacy Scale

1

98


0.13 [-0.27, 0.53]

3 3

145 145


0.19 [-1.92, 2.31] 0.19 [-1.92, 2.31]

1 1

98 98


0.30 [-0.71, 1.31] 0.30 [-0.71, 1.31]

2 1 1

117 19 98


-0.04 [-0.83, 0.75] -0.56 [-1.49, 0.36] 0.27 [-0.14, 0.67]

1 1

28 28


6.08 [-1.57, 13.73] 6.08 [-1.57, 13.73]

2 1

159 98


0.13 [-0.19, 0.44] 0.12 [-0.28, 0.53]

1

61


0.13 [-0.38, 0.64]


79

Analysis 1.1. Comparison 1 Cognitive training versus any control, Outcome 1 Attention. Review:

Neuropsychological rehabilitation for multiple sclerosis

Comparison: 1 Cognitive training versus any control Outcome: 1 Attention

Study or subgroup

Intervention N

Std. Mean Difference

Control Mean(SD)

N

Mean(SD)

Weight

IV,Fixed,95% CI

Std. Mean Difference IV,Fixed,95% CI

1 Paced Auditory Serial Addition Test 2’ Mattioli 2010

10

23.4 (13.7)

10

6.5 (9.9)

2.8 %

1.35 [ 0.36, 2.35 ]

Solari 2004

40

22.6 (10.1)

37

25.9 (16.4)

13.6 %

-0.24 [ -0.69, 0.21 ]

16.4 %

0.03 [ -0.38, 0.44 ]

Subtotal (95% CI)

50

47

Heterogeneity: Chi2 = 8.23, df = 1 (P = 0.004); I2 =88% Test for overall effect: Z = 0.14 (P = 0.89) 2 Paced Auditory Serial Addition Test 3’ Cerasa 2012

12

41.23 (12.7)

11

41 (8.79)

4.1 %

0.02 [ -0.80, 0.84 ]

Mattioli 2010

10

34.7 (14.8)

10

15.4 (18.4)

3.0 %

1.11 [ 0.15, 2.06 ]

Vogt 2008

15

50.4 (7.91)

15

48.53 (11.1)

5.3 %

0.19 [ -0.53, 0.91 ]

12.4 %

0.35 [ -0.12, 0.82 ]

Subtotal (95% CI)

37

36

Heterogeneity: Chi2 = 3.22, df = 2 (P = 0.20); I2 =38% Test for overall effect: Z = 1.48 (P = 0.14) 3 Symbol Digit Modalities Test Cerasa 2012

12

38.69 (9.9)

11

37.3 (8.45)

4.1 %

0.15 [ -0.67, 0.96 ]

Mattioli 2010

10

36.3 (9.9)

10

37.9 (17.7)

3.6 %

-0.11 [ -0.98, 0.77 ]

Solari 2004

40

30 (6.5)

37

29.1 (7.6)

13.7 %

0.13 [ -0.32, 0.57 ]

Vogt 2008

15

53.87 (14.78)

15

58.67 (19.19)

5.3 %

-0.27 [ -0.99, 0.45 ]

26.6 %

0.02 [ -0.30, 0.34 ]

8.2 %

-0.05 [ -0.63, 0.53 ]

8.2 %

-0.05 [ -0.63, 0.53 ]

8.2 %

-0.05 [ -0.63, 0.53 ]

8.2 %

-0.05 [ -0.63, 0.53 ]

Subtotal (95% CI)

77

73

Heterogeneity: Chi2 = 1.02, df = 3 (P = 0.80); I2 =0.0% Test for overall effect: Z = 0.11 (P = 0.91) 4 Divided attention Shatil 2010

Subtotal (95% CI)

22

2.37 (0.78)

22

24

2.41 (0.72)

24

Heterogeneity: not applicable Test for overall effect: Z = 0.18 (P = 0.86) 5 Avoiding distractions Shatil 2010

Subtotal (95% CI)

22

22

-0.7 (0.47)

24

-0.67 (0.69)

24

Heterogeneity: not applicable

-2

-1

Favours control

0

1

2

Favours intervention

(Continued . . . )


80

(. . . Study or subgroup

Intervention N


Control Mean(SD)

N

Mean(SD)

-0.16 (0.62)

24

-0.3 (0.59)

Weight

IV,Fixed,95% CI

Continued) Std. Mean Difference

IV,Fixed,95% CI

Test for overall effect: Z = 0.17 (P = 0.87) 6 Inhibition Shatil 2010

Subtotal (95% CI)

22

22

8.1 %

0.23 [ -0.35, 0.81 ]

8.1 %

0.23 [ -0.35, 0.81 ]

8.2 %

-0.14 [ -0.72, 0.44 ]

8.2 %

-0.14 [ -0.72, 0.44 ]

4.0 %

-0.28 [ -1.10, 0.54 ]

4.0 %

-0.28 [ -1.10, 0.54 ]

4.1 %

0.01 [ -0.81, 0.82 ]

4.1 %

0.01 [ -0.81, 0.82 ]

3.9 %

0.54 [ -0.29, 1.38 ]

11

3.9 %

0.54 [ -0.29, 1.38 ]

285

100.0 %

0.06 [ -0.10, 0.23 ]

24

Heterogeneity: not applicable Test for overall effect: Z = 0.77 (P = 0.44) 7 Shifting attention Shatil 2010

Subtotal (95% CI)

22

0.37 (0.91)

22

24

0.48 (0.62)

24

Heterogeneity: not applicable Test for overall effect: Z = 0.47 (P = 0.64) 8 Trail Making A Cerasa 2012

Subtotal (95% CI)

12

-44.83 (13.1)

12

11

-40.9 (13.94)

11

Heterogeneity: not applicable Test for overall effect: Z = 0.67 (P = 0.50) 9 Trail Making B Cerasa 2012

Subtotal (95% CI)

12

-120.9 (37.9)

12

11

-121.1 (37.4)

11

Heterogeneity: not applicable Test for overall effect: Z = 0.01 (P = 0.99) 10 Stroop/colour-word Cerasa 2012

Subtotal (95% CI)

12

19.41 (5.14)

12

11

16.5 (5.22)

Heterogeneity: not applicable Test for overall effect: Z = 1.27 (P = 0.20)

Total (95% CI)

288

Heterogeneity: Chi2 = 17.09, df = 15 (P = 0.31); I2 =12% Test for overall effect: Z = 0.74 (P = 0.46) Test for subgroup differences: Chi2 = 4.61, df = 9 (P = 0.87), I2 =0.0%

-2

-1

Favours control


0

1

2


81

Analysis 1.2. Comparison 1 Cognitive training versus any control, Outcome 2 Information processing speed. Review:


Comparison: 1 Cognitive training versus any control Outcome: 2 Information processing speed

Study or subgroup

Intervention


Control

N

Mean(SD)

N

Mean(SD)

-1.2 (0.3)

40

-1.5 (0.4)

Weight

IV,Random,95% CI

Std. Mean Difference IV,Random,95% CI

1 Signal detection time/reaction time(s) Mendozzi 1998

Subtotal (95% CI)

20

20

40

24.4 %

0.80 [ 0.24, 1.36 ]

24.4 %

0.80 [ 0.24, 1.36 ]

23.7 %

0.17 [ -0.41, 0.75 ]

23.7 %

0.17 [ -0.41, 0.75 ]

20.0 %

-0.02 [ -0.74, 0.69 ]

Heterogeneity: not applicable Test for overall effect: Z = 2.82 (P = 0.0049) 2 Response time Shatil 2010

Subtotal (95% CI)

22

-0.39 (0.74)

22

24

-0.51 (0.67)

24

Heterogeneity: not applicable Test for overall effect: Z = 0.57 (P = 0.57) 3 2-back (reaction time) Vogt 2008

Subtotal (95% CI)

15

-767.77 (272.31)

15

15

-762.07 (257.49)

15

20.0 % -0.02 [ -0.74, 0.69 ]

Heterogeneity: not applicable Test for overall effect: Z = 0.06 (P = 0.95) 4 Test of Everyday Attention (median for auditory stimulus) Mattioli 2010

Subtotal (95% CI)

10

-711.8 (160.6)

10

10

15.6 %

-605.5 (96.5)

-0.77 [ -1.68, 0.15 ]

15.6 % -0.77 [ -1.68, 0.15 ]

10

Heterogeneity: not applicable Test for overall effect: Z = 1.64 (P = 0.10) 5 Test of Everyday Attention (median for visual stimulus) Mattioli 2010

Subtotal (95% CI)

10

10

-961.2 (195.2)

10

-1003.5 (191.2)

16.3 %

0.21 [ -0.67, 1.09 ]

10

16.3 %

0.21 [ -0.67, 1.09 ]

99

100.0 %

0.15 [ -0.33, 0.62 ]


Total (95% CI)

77

Heterogeneity: Tau2 = 0.16; Chi2 = 9.11, df = 4 (P = 0.06); I2 =56% Test for overall effect: Z = 0.60 (P = 0.55) Test for subgroup differences: Chi2 = 9.11, df = 4 (P = 0.06), I2 =56%

-1

-0.5

Favours control


0

0.5

1


82

Analysis 1.3. Comparison 1 Cognitive training versus any control, Outcome 3 Memory span. Review:


Comparison: 1 Cognitive training versus any control Outcome: 3 Memory span

Study or subgroup

Intervention


Control

Weight

IV,Fixed,95% CI


N

Mean(SD)

N

Mean(SD)

IV,Fixed,95% CI

Mendozzi 1998

20

5 (0.8)

40

4.4 (0.8)

38.1 %

0.74 [ 0.19, 1.29 ]

Vogt 2008

15

9.2 (1.93)

15

8.8 (1.52)

22.6 %

0.22 [ -0.49, 0.94 ]

60.7 %

0.55 [ 0.11, 0.99 ]

39.3 %

0.52 [ -0.02, 1.07 ]

40

39.3 %

0.52 [ -0.02, 1.07 ]

95

100.0 %

0.54 [ 0.20, 0.88 ]

1 Corsi blocks forward

Subtotal (95% CI)

35

55

Heterogeneity: Chi2 = 1.24, df = 1 (P = 0.26); I2 =20% Test for overall effect: Z = 2.45 (P = 0.014) 2 Digit span forward Mendozzi 1998

Subtotal (95% CI)

20

6.1 (1.2)

20

40

5.5 (1.1)


Total (95% CI)

55

Heterogeneity: Chi2 = 1.25, df = 2 (P = 0.54); I2 =0.0% Test for overall effect: Z = 3.08 (P = 0.0020) Test for subgroup differences: Chi2 = 0.01, df = 1 (P = 0.94), I2 =0.0%

-1

-0.5

Favours control


0

0.5

1


83

Analysis 1.4. Comparison 1 Cognitive training versus any control, Outcome 4 Working memory. Review:


Comparison: 1 Cognitive training versus any control Outcome: 4 Working memory

Study or subgroup

Intervention


Control

N

Mean(SD)

N

Mean(SD)

15

55.07 (4.02)

15

55.27 (3.92)

Weight

IV,Fixed,95% CI


1 2-back (numbers correct) Vogt 2008

Subtotal (95% CI)

15

15

10.9 %

-0.05 [ -0.76, 0.67 ]

10.9 %

-0.05 [ -0.76, 0.67 ]

Heterogeneity: not applicable Test for overall effect: Z = 0.13 (P = 0.89) 2 Digit span backward Mendozzi 1998

20

4 (1.3)

40

3.8 (1)

19.3 %

0.18 [ -0.36, 0.72 ]

Vogt 2008

15

7.87 (2.38)

15

6.4 (1.99)

10.3 %

0.65 [ -0.09, 1.39 ]

29.6 %

0.34 [ -0.09, 0.78 ]

10.7 %

0.38 [ -0.34, 1.10 ]

10.7 %

0.38 [ -0.34, 1.10 ]

15.9 %

0.59 [ 0.00, 1.19 ]

15.9 %

0.59 [ 0.00, 1.19 ]

16.1 %

0.52 [ -0.07, 1.11 ]

16.1 %

0.52 [ -0.07, 1.11 ]

16.7 %

0.08 [ -0.50, 0.66 ]

24

16.7 %

0.08 [ -0.50, 0.66 ]

157

100.0 %

0.33 [ 0.09, 0.57 ]

Subtotal (95% CI)

35

55

Heterogeneity: Chi2 = 1.04, df = 1 (P = 0.31); I2 =3% Test for overall effect: Z = 1.55 (P = 0.12) 3 Corsi blocks backward Vogt 2008

Subtotal (95% CI)

15

8.87 (2.03)

15

15

8.13 (1.76)

15

Heterogeneity: not applicable Test for overall effect: Z = 1.03 (P = 0.30) 4 Verbal auditory working memory Shatil 2010

Subtotal (95% CI)

22

1.09 (0.81)

22

24

0.53 (1.02)

24

Heterogeneity: not applicable Test for overall effect: Z = 1.97 (P = 0.049) 5 Visual working memory Shatil 2010

Subtotal (95% CI)

22

1.15 (0.84)

22

24

0.65 (1.03)

24

Heterogeneity: not applicable Test for overall effect: Z = 1.73 (P = 0.083) 6 Auditory working memory Shatil 2010

Subtotal (95% CI)

22

22

0.69 (0.78)

24

0.62 (0.87)


Total (95% CI)

131

-1

-0.5

Favours control

0

0.5

1


(Continued . . . )


84


Intervention

Control

N Heterogeneity:

Chi2


Mean(SD)

= 4.00, df = 6 (P = 0.68);

I2

N

Mean(SD)

Weight

IV,Fixed,95% CI


IV,Fixed,95% CI

=0.0%

Test for overall effect: Z = 2.73 (P = 0.0063) Test for subgroup differences: Chi2 = 2.97, df = 5 (P = 0.71), I2 =0.0%

-1

-0.5

0

Favours control

0.5

1


Analysis 1.5. Comparison 1 Cognitive training versus any control, Outcome 5 Immediate verbal memory. Review:


Comparison: 1 Cognitive training versus any control Outcome: 5 Immediate verbal memory

Study or subgroup

Intervention N


Control Mean(SD)

N

Mean(SD)

Weight

IV,Fixed,95% CI


1 Buschke Selective Reminding Test (consistent long-term retrieval) Cerasa 2012

12

24.86 (11.05)

11

17.1 (7.3)

6.4 %

0.79 [ -0.06, 1.65 ]

Mattioli 2010

10

22 (15.5)

10

17.6 (12.4)

6.0 %

0.30 [ -0.58, 1.18 ]

Solari 2004

40

30 (38.7)

37

26.1 (29.3)

23.3 %

0.11 [ -0.34, 0.56 ]

35.6 %

0.26 [ -0.10, 0.63 ]

16.2 %

-0.08 [ -0.62, 0.46 ]

16.2 %

-0.08 [ -0.62, 0.46 ]

16.0 %

0.29 [ -0.25, 0.83 ]

16.0 %

0.29 [ -0.25, 0.83 ]

Subtotal (95% CI)

62

58

Heterogeneity: Chi2 = 1.91, df = 2 (P = 0.39); I2 =0.0% Test for overall effect: Z = 1.44 (P = 0.15) 2 Wechsler Memory Scale/logical memory Mendozzi 1998

Subtotal (95% CI)

20

14 (6.8)

20

40

14.5 (5.9)

40

Heterogeneity: not applicable Test for overall effect: Z = 0.29 (P = 0.77) 3 Wechsler Memory Scale/verbal paired associations (easy) Mendozzi 1998

Subtotal (95% CI)

20

20

15.9 (2.2)

40

15.3 (2)

40 -1

-0.5

Favours control

0

0.5

1


(Continued . . . )


85


Intervention N


Control Mean(SD)

N

Mean(SD)

40

4.3 (3)

Weight

IV,Fixed,95% CI


IV,Fixed,95% CI

Heterogeneity: not applicable Test for overall effect: Z = 1.04 (P = 0.30) 4 Wechsler Memory Scale/verbal paired associations (hard) Mendozzi 1998

Subtotal (95% CI)

20

5.3 (3.4)

20

16.0 %

0.31 [ -0.23, 0.85 ]

16.0 %

0.31 [ -0.23, 0.85 ]

16.2 %

0.11 [ -0.43, 0.65 ]

40

16.2 %

0.11 [ -0.43, 0.65 ]

218

100.0 %

0.20 [ -0.02, 0.41 ]

40

Heterogeneity: not applicable Test for overall effect: Z = 1.14 (P = 0.25) 5 Memory scale of the Luria-Nebraska Neuropsychological Battery Mendozzi 1998

Subtotal (95% CI)

20

-57.7 (5.8)

20

40

-58.4 (6.4)


Total (95% CI)

142


-1

-0.5

Favours control


0

0.5

1


86

Analysis 1.6. Comparison 1 Cognitive training versus any control, Outcome 6 Immediate visual memory. Review:


Comparison: 1 Cognitive training versus any control Outcome: 6 Immediate visual memory

Study or subgroup

Intervention


Control

N

Mean(SD)

N

Mean(SD)

20

76.5 (10.2)

40

71.9 (13.4)

Weight

IV,Random,95% CI


1 Recognition memory (Corsi) Mendozzi 1998

Subtotal (95% CI)

20

40

19.1 %

0.36 [ -0.18, 0.91 ]

19.1 %

0.36 [ -0.18, 0.91 ]

18.7 %

0.79 [ 0.23, 1.34 ]

18.7 %

0.79 [ 0.23, 1.34 ]

Heterogeneity: not applicable Test for overall effect: Z = 1.32 (P = 0.19) 2 Wechsler Memory Scale/visual reproduction Mendozzi 1998

Subtotal (95% CI)

20

8.2 (3.3)

20

40

5.3 (3.8)

40

Heterogeneity: not applicable Test for overall effect: Z = 2.77 (P = 0.0056) 3 10/36 Spatial Recall Test (immediate recall) Cerasa 2012

12

18.42 (6.22)

11

24.3 (3.99)

12.8 %

-1.07 [ -1.96, -0.19 ]

Mattioli 2010

10

16.9 (4.6)

10

14.7 (4.5)

12.7 %

0.46 [ -0.43, 1.35 ]

Solari 2004

40

16.7 (6.3)

37

17.1 (9.1)

21.0 %

-0.05 [ -0.50, 0.40 ]

46.5 %

-0.20 [ -0.94, 0.54 ]

15.7 %

0.06 [ -0.66, 0.77 ]

15

15.7 %

0.06 [ -0.66, 0.77 ]

153

100.0 %

0.14 [ -0.30, 0.58 ]

Subtotal (95% CI)

62

58

Heterogeneity: Tau2 = 0.29; Chi2 = 6.18, df = 2 (P = 0.05); I2 =68% Test for overall effect: Z = 0.53 (P = 0.60) 4 Faces Symbol Test Vogt 2008

Subtotal (95% CI)

15

15

2.54 (0.73)

15

2.49 (0.91)


Total (95% CI)

117


-1

-0.5

Favours control


0

0.5

1


87

Analysis 1.7. Comparison 1 Cognitive training versus any control, Outcome 7 Delayed memory. Review:


Comparison: 1 Cognitive training versus any control Outcome: 7 Delayed memory

Study or subgroup

Intervention N


Control Mean(SD)

N

Mean(SD)

Weight

IV,Fixed,95% CI


1 Buschke Selective Reminding Test (delayed recall) Cerasa 2012

12

7.11 (2.93)

11

6.2 (3.02)

9.7 %

0.30 [ -0.53, 1.12 ]

Mattioli 2010

10

6.6 (2.7)

10

6 (2.5)

8.5 %

0.22 [ -0.66, 1.10 ]

Solari 2004

40

5.9 (2.3)

37

7 (5.6)

32.6 %

-0.26 [ -0.71, 0.19 ]

50.8 %

-0.07 [ -0.43, 0.29 ]

Subtotal (95% CI)

62

58

Heterogeneity: Chi2 = 1.85, df = 2 (P = 0.40); I2 =0.0% Test for overall effect: Z = 0.39 (P = 0.69) 2 10/36 Spatial Recall Test (delayed recall) Cerasa 2012

12

5.58 (2.47)

11

8.3 (1.89)

8.1 %

-1.18 [ -2.09, -0.28 ]

Mattioli 2010

10

5.7 (2.6)

10

4.9 (2.4)

8.4 %

0.31 [ -0.58, 1.19 ]

Solari 2004

40

5.9 (2.3)

37

7 (5.6)

32.6 %

-0.26 [ -0.71, 0.19 ]

58

49.2 %

-0.31 [ -0.68, 0.05 ]

116

100.0 %

-0.19 [ -0.45, 0.07 ]

Subtotal (95% CI)

62

Heterogeneity: Chi2 = 5.55, df = 2 (P = 0.06); I2 =64% Test for overall effect: Z = 1.68 (P = 0.092)

Total (95% CI)

124


-1

-0.5

Favours control


0

0.5

1


88

Analysis 1.8. Comparison 1 Cognitive training versus any control, Outcome 8 Executive functions. Review:


Comparison: 1 Cognitive training versus any control Outcome: 8 Executive functions

Study or subgroup

Intervention N


Control Mean(SD)

N

Mean(SD)

-21.2 (9.1)

10

-43.5 (24.4)

Weight

IV,Fixed,95% CI


1 Wisconsin Card Sorting Test (total errors) Mattioli 2010

Subtotal (95% CI)

10

10

15.4 %

1.16 [ 0.20, 2.12 ]

15.4 %

1.16 [ 0.20, 2.12 ]

42.6 %

0.08 [ -0.50, 0.66 ]

42.6 %

0.08 [ -0.50, 0.66 ]

42.0 %

0.33 [ -0.25, 0.91 ]

24

42.0 %

0.33 [ -0.25, 0.91 ]

58

100.0 %

0.35 [ -0.03, 0.73 ]

10

Heterogeneity: not applicable Test for overall effect: Z = 2.36 (P = 0.018) 2 Planning Shatil 2010

Subtotal (95% CI)

22

0.07 (0.77)

22

24

0 (0.9)

24

Heterogeneity: not applicable Test for overall effect: Z = 0.28 (P = 0.78) 3 Time estimation Shatil 2010

Subtotal (95% CI)

22

0.62 (0.61)

22

24

0.34 (1)


Total (95% CI)

54

Heterogeneity: Chi2 = 3.54, df = 2 (P = 0.17); I2 =44% Test for overall effect: Z = 1.82 (P = 0.068) Test for subgroup differences: Chi2 = 3.54, df = 2 (P = 0.17), I2 =44%

-1

-0.5

Favours control


0

0.5

1


89

Analysis 1.9. Comparison 1 Cognitive training versus any control, Outcome 9 Verbal functions. Review:


Comparison: 1 Cognitive training versus any control Outcome: 9 Verbal functions

Study or subgroup

Intervention


Control

N

Mean(SD)

N

Mean(SD)

22

0.68 (0.56)

24

0.54 (0.85)

Weight

IV,Fixed,95% CI


1 Naming Shatil 2010

Subtotal (95% CI)

22

24

24.8 %

0.19 [ -0.39, 0.77 ]

24.8 %

0.19 [ -0.39, 0.77 ]

Heterogeneity: not applicable Test for overall effect: Z = 0.64 (P = 0.52) 2 Controlled Oral Word Association Test (phonemic) Mattioli 2010

10

35.8 (10.3)

10

29.3 (12.6)

10.4 %

0.54 [ -0.36, 1.44 ]

Solari 2004

40

22.9 (9.6)

37

21.8 (7)

41.7 %

0.13 [ -0.32, 0.58 ]

52.1 %

0.21 [ -0.19, 0.61 ]

Subtotal (95% CI)

50

47

Heterogeneity: Chi2 = 0.65, df = 1 (P = 0.42); I2 =0.0% Test for overall effect: Z = 1.03 (P = 0.30) 3 Controlled Oral Word Association Test (semantic) Cerasa 2012

12

20.8 (5.96)

11

20.6 (5.59)

12.5 %

0.03 [ -0.78, 0.85 ]

Mattioli 2010

10

39.6 (12.6)

10

35.2 (8.3)

10.6 %

0.39 [ -0.49, 1.28 ]

21

23.1 %

0.20 [ -0.40, 0.80 ]

92

100.0 %

0.20 [ -0.09, 0.49 ]

Subtotal (95% CI)

22

Heterogeneity: Chi2 = 0.35, df = 1 (P = 0.56); I2 =0.0% Test for overall effect: Z = 0.65 (P = 0.52)

Total (95% CI)

94


-1

-0.5

Favours control


0

0.5

1


90

Analysis 1.10. Comparison 1 Cognitive training versus any control, Outcome 10 Depression. Review:


Comparison: 1 Cognitive training versus any control Outcome: 10 Depression

Study or subgroup

Intervention N


Control Mean(SD)

N

Mean(SD)

10

-15.5 (8.8)

Weight

IV,Random,95% CI


1 Montgomery-Asberg Depression Rating Scale Mattioli 2010

Subtotal (95% CI)

10

-6 (5.1)

10

14.4 %

1.27 [ 0.29, 2.24 ]

14.4 %

1.27 [ 0.29, 2.24 ]

26.4 %

0.0 [ -0.45, 0.45 ]

26.4 %

0.0 [ -0.45, 0.45 ]

19.7 %

0.05 [ -0.66, 0.77 ]

19.7 %

0.05 [ -0.66, 0.77 ]

22.9 %

-0.31 [ -0.89, 0.27 ]

22.9 %

-0.31 [ -0.89, 0.27 ]

16.6 %

0.85 [ -0.01, 1.71 ]

11

16.6 %

0.85 [ -0.01, 1.71 ]

97

100.0 %

0.26 [ -0.23, 0.75 ]

10

Heterogeneity: not applicable Test for overall effect: Z = 2.53 (P = 0.011) 2 Chicago Mood Depression Inventory Solari 2004

Subtotal (95% CI)

40

-28.4 (11.1)

40

37

-28.4 (10.3)

37

Heterogeneity: not applicable Test for overall effect: Z = 0.0 (P = 1.0) 3 Allgemeine Depressionskala Vogt 2008

Subtotal (95% CI)

15

-12.21 (12.2)

15

15

-12.86 (10.98)

15

Heterogeneity: not applicable Test for overall effect: Z = 0.15 (P = 0.88) 4 Zung Depression Scale Shatil 2010

Subtotal (95% CI)

22

-61.36 (6.76)

22

24

-59.27 (6.49)

24

Heterogeneity: not applicable Test for overall effect: Z = 1.04 (P = 0.30) 5 Beck Depression Inventory II Cerasa 2012

Subtotal (95% CI)

12

12

-4.33 (3.17)

11

-12.8 (13.5)


Total (95% CI)

99


-2

-1

Favours control


0

1

2


91

Analysis 1.11. Comparison 1 Cognitive training versus any control, Outcome 11 Quality of life. Review:


Comparison: 1 Cognitive training versus any control Outcome: 11 Quality of life

Study or subgroup

Intervention


Control

N

Mean(SD)

N

Mean(SD)

10

185.2 (36.8)

10

167.3 (36.7)

Weight

IV,Fixed,95% CI


1 Multiple Sclerosis Quality of Life Mattioli 2010

Subtotal (95% CI)

10

15.5 %

0.47 [ -0.42, 1.36 ]

15.5 %

0.47 [ -0.42, 1.36 ]

24.0 %

-0.13 [ -0.84, 0.59 ]

24.0 %

-0.13 [ -0.84, 0.59 ]

60.5 %

0.36 [ -0.09, 0.81 ]

37

60.5 %

0.36 [ -0.09, 0.81 ]

62

100.0 %

0.26 [ -0.09, 0.61 ]

10

Heterogeneity: not applicable Test for overall effect: Z = 1.03 (P = 0.30) 2 Functional Assessment of MS Vogt 2008

Subtotal (95% CI)

15

118.6 (34.08)

15

15

122.93 (32.77)

15

Heterogeneity: not applicable Test for overall effect: Z = 0.34 (P = 0.73) 3 54-item MS Quality of Life Questionnaire/mental health composite Solari 2004

Subtotal (95% CI)

40

72.1 (50.9)

40

37

56.8 (29.8)


Total (95% CI)

65


-1

-0.5

Favours control


0

0.5

1


92

Analysis 1.12. Comparison 1 Cognitive training versus any control, Outcome 12 Fatigue. Review:


Comparison: 1 Cognitive training versus any control Outcome: 12 Fatigue

Study or subgroup

Intervention N


Control Mean(SD)

N

Mean(SD)

15

-65.06 (16.68)

Weight

IV,Fixed,95% CI


1 Fatigue Scale for Motor and Cognitive functions Vogt 2008

Subtotal (95% CI)

15

-61.73 (19.08)

15

23.7 %

0.18 [ -0.54, 0.90 ]

23.7 %

0.18 [ -0.54, 0.90 ]

23.6 %

0.19 [ -0.53, 0.91 ]

23.6 %

0.19 [ -0.53, 0.91 ]

36.1 %

-0.25 [ -0.83, 0.33 ]

36.1 %

-0.25 [ -0.83, 0.33 ]

16.6 %

0.79 [ -0.07, 1.64 ]

11

16.6 %

0.79 [ -0.07, 1.64 ]

65

100.0 %

0.13 [ -0.22, 0.48 ]

15

Heterogeneity: not applicable Test for overall effect: Z = 0.49 (P = 0.62) 2 Modified Fatigue Impact Scale Vogt 2008

Subtotal (95% CI)

15

-34.13 (17.34)

15

15

-37.53 (17.29)

15

Heterogeneity: not applicable Test for overall effect: Z = 0.52 (P = 0.60) 3 Fatigue Severity Scale/raw score Shatil 2010

Subtotal (95% CI)

22

-44 (14.51)

22

24

-39.83 (17.85)

24

Heterogeneity: not applicable Test for overall effect: Z = 0.85 (P = 0.40) 4 Fatigue Severity Scale/delta score Cerasa 2012

Subtotal (95% CI)

12

-2.83 (1.63)

12

11

-4.22 (1.77)


Total (95% CI)

64


-1

-0.5

Favours control


0

0.5

1


93

Analysis 1.13. Comparison 1 Cognitive training versus any control, Outcome 13 Anxiety. Review:


Comparison: 1 Cognitive training versus any control Outcome: 13 Anxiety

Study or subgroup

Intervention

Mean Difference

Control

N

Mean(SD)

N

Mean(SD)

12

-36.6 (8.9)

11

-41 (11.1)

Weight

IV,Fixed,95% CI

Mean Difference IV,Fixed,95% CI

1 State Trait Anxiety Inventory Cerasa 2012

Total (95% CI)

12

11

100.0 %

4.40 [ -3.87, 12.67 ]

100.0 %

4.40 [ -3.87, 12.67 ]

Heterogeneity: not applicable Test for overall effect: Z = 1.04 (P = 0.30) Test for subgroup differences: Not applicable

-20

-10

Favours control


0

10

20


94

Analysis 2.1. Comparison 2 Cognitive training versus any control (longitudinal follow-up), Outcome 1 Attention. Review:


Comparison: 2 Cognitive training versus any control (longitudinal follow-up) Outcome: 1 Attention

Study or subgroup

Intervention N


Control Mean(SD)

N

Mean(SD)

21.2 (9)

37

27 (19.7)

Weight

IV,Fixed,95% CI


1 Paced Auditory Serial Addition Test 2’ Solari 2004

Subtotal (95% CI)

40

40

49.5 %

-0.38 [ -0.83, 0.07 ]

49.5 %

-0.38 [ -0.83, 0.07 ]

50.5 %

-0.06 [ -0.51, 0.39 ]

37

50.5 %

-0.06 [ -0.51, 0.39 ]

74

100.0 %

-0.22 [ -0.54, 0.10 ]

37

Heterogeneity: not applicable Test for overall effect: Z = 1.65 (P = 0.099) 2 Symbol Digit Modalities Test Solari 2004

Subtotal (95% CI)

40

30.6 (7.1)

40

37

31.1 (9.6)


Total (95% CI)

80


-1

-0.5

Favours control


0

0.5

1


95

Analysis 2.2. Comparison 2 Cognitive training versus any control (longitudinal follow-up), Outcome 2 Immediate verbal memory. Review:


Comparison: 2 Cognitive training versus any control (longitudinal follow-up) Outcome: 2 Immediate verbal memory

Study or subgroup

Intervention N

Mean Difference

Control Mean(SD)

N

Mean(SD)

Weight

IV,Fixed,95% CI


1 Buschke Selective Reminding Test/consistent long-term retrieval Solari 2004

40

Total (95% CI)

32.8 (39.6)

40

37

33.1 (38.6)

37

100.0 %

-0.30 [ -17.77, 17.17 ]

100.0 %

-0.30 [ -17.77, 17.17 ]


-20

-10

0

10

Favours control

20


Analysis 2.3. Comparison 2 Cognitive training versus any control (longitudinal follow-up), Outcome 3 Immediate visual memory. Review:


Comparison: 2 Cognitive training versus any control (longitudinal follow-up) Outcome: 3 Immediate visual memory

Study or subgroup

Intervention N

Mean Difference

Control Mean(SD)

N

Mean(SD)

15.8 (7.2)

37

17 (9)

Weight

IV,Fixed,95% CI


1 10/36 Spatial Recall Test/immediate recall Solari 2004

Total (95% CI)

40

40

37

100.0 %

-1.20 [ -4.86, 2.46 ]

100.0 %

-1.20 [ -4.86, 2.46 ]


-10

-5

Favours control


0

5

10


96

Analysis 2.4. Comparison 2 Cognitive training versus any control (longitudinal follow-up), Outcome 4 Delayed memory. Review:


Comparison: 2 Cognitive training versus any control (longitudinal follow-up) Outcome: 4 Delayed memory

Study or subgroup

Intervention N


Control Mean(SD)

N

Mean(SD)

37

7.5 (5.1)

Weight

IV,Fixed,95% CI


1 Buschke Selective Reminding Test/delayed recall Solari 2004

Subtotal (95% CI)

40

6.3 (2.3)

40

50.0 %

-0.30 [ -0.75, 0.15 ]

50.0 %

-0.30 [ -0.75, 0.15 ]

50.0 %

-0.30 [ -0.75, 0.15 ]

37

50.0 %

-0.30 [ -0.75, 0.15 ]

74

100.0 %

-0.30 [ -0.62, 0.02 ]

37

Heterogeneity: not applicable Test for overall effect: Z = 1.33 (P = 0.18) 2 10/36 Spatial Recall Test/delayed recall Solari 2004

Subtotal (95% CI)

40

5.5 (3.1)

40

37

6.9 (5.8)


Total (95% CI)

80


-2

-1

Favours control


0

1

2


97

Analysis 2.5. Comparison 2 Cognitive training versus any control (longitudinal follow-up), Outcome 5 Verbal functions. Review:


Comparison: 2 Cognitive training versus any control (longitudinal follow-up) Outcome: 5 Verbal functions

Study or subgroup

Intervention N

Mean Difference

Control Mean(SD)

N

Mean(SD)

37

26.3 (7.7)

Weight

IV,Fixed,95% CI


1 Controlled Oral Word Association Test (phonemic) Solari 2004

40

Total (95% CI)

28.4 (10)

40

37

100.0 %

2.10 [ -1.87, 6.07 ]

100.0 %

2.10 [ -1.87, 6.07 ]


-4

-2

0

Favours control

2

4


Analysis 2.6. Comparison 2 Cognitive training versus any control (longitudinal follow-up), Outcome 6 Depression. Review:


Comparison: 2 Cognitive training versus any control (longitudinal follow-up) Outcome: 6 Depression

Study or subgroup

Intervention N

Mean Difference

Control Mean(SD)

N

Mean(SD)

-27.3 (11.1)

37

-26.9 (10.2)

Weight

IV,Fixed,95% CI


1 Chicago Mood Depression Inventory Solari 2004

Total (95% CI)

40

40

37

100.0 %

-0.40 [ -5.16, 4.36 ]

100.0 %

-0.40 [ -5.16, 4.36 ]


-10

-5

Favours control


0

5

10


98

Analysis 2.7. Comparison 2 Cognitive training versus any control (longitudinal follow-up), Outcome 7 Quality of life. Review:


Comparison: 2 Cognitive training versus any control (longitudinal follow-up) Outcome: 7 Quality of life

Study or subgroup

Intervention N

Mean Difference

Control Mean(SD)

N

Mean(SD)

Weight

IV,Fixed,95% CI


1 54-item MS Quality of Life Questionnaire/mental health composite Solari 2004

Total (95% CI)

40

60.8 (32.4)

40

37

100.0 %

-9.40 [ -28.38, 9.58 ]

100.0 %

-9.40 [ -28.38, 9.58 ]

70.2 (50)

37


-100

-50

Favours control


0

50

100


99

Analysis 3.1. Comparison 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control, Outcome 1 Attention. Review:


Comparison: 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control Outcome: 1 Attention

Study or subgroup

Intervention N


Control Mean(SD)

N

Mean(SD)

47.1 (19.9)

9

47.5 (18.6)

Weight

IV,Fixed,95% CI


1 Sustained attention test/correct answers Tesar 2005

Subtotal (95% CI)

10

10

9

2.2 %

-0.02 [ -0.92, 0.88 ]

2.2 %

-0.02 [ -0.92, 0.88 ]

Heterogeneity: not applicable Test for overall effect: Z = 0.04 (P = 0.97) 2 Paced Auditory Serial Addition Test 3’/raw score Mäntynen 2013

58

43.8 (14.3)

40

41 (12.5)

10.9 %

0.20 [ -0.20, 0.61 ]

Stuifbergen 2012

34

45.2 (11.2)

27

46.7 (11.2)

7.0 %

-0.13 [ -0.64, 0.37 ]

17.9 %

0.07 [ -0.24, 0.39 ]

Subtotal (95% CI)

92

67

Heterogeneity: Chi2 = 1.04, df = 1 (P = 0.31); I2 =4% Test for overall effect: Z = 0.46 (P = 0.65) 3 Paced Auditory Serial Addition Test 2’/raw score Mäntynen 2013

58

32.3 (12.7)

40

27.5 (10)

10.7 %

0.41 [ 0.00, 0.81 ]

Stuifbergen 2012

34

34 (9.3)

27

34.9 (11.5)

7.0 %

-0.09 [ -0.59, 0.42 ]

17.7 %

0.21 [ -0.10, 0.53 ]

4.7 %

0.01 [ -0.61, 0.62 ]

4.7 %

0.01 [ -0.61, 0.62 ]

4.7 %

0.23 [ -0.39, 0.85 ]

4.7 %

0.23 [ -0.39, 0.85 ]

Subtotal (95% CI)

92

67

Heterogeneity: Chi2 = 2.22, df = 1 (P = 0.14); I2 =55% Test for overall effect: Z = 1.32 (P = 0.19) 4 Paced Auditory Serial Addition Test 3’/z-score Hildebrandt 2007

17

Subtotal (95% CI)

17

0.017 (0.83)

25

0.01 (1.09)

25

Heterogeneity: not applicable Test for overall effect: Z = 0.02 (P = 0.98) 5 Test battery of attention/alertness without cueing Hildebrandt 2007

17

Subtotal (95% CI)

17

248 (85)

25

233 (45)

25

Heterogeneity: not applicable Test for overall effect: Z = 0.73 (P = 0.47) 6 Symbol Digit Modalities Test Mäntynen 2013

58

49.8 (10.2)

40

47.5 (8.4)

10.9 %

0.24 [ -0.16, 0.64 ]

Stuifbergen 2012

34

49.6 (11.1)

27

48.1 (14)

7.0 %

0.12 [ -0.39, 0.62 ]

-0.5

-0.25

Favours control

0

0.25

0.5


(Continued . . . )


100


Intervention

Control

N

Subtotal (95% CI)


Mean(SD)

92

N

Mean(SD)

Weight

IV,Fixed,95% CI


IV,Fixed,95% CI

67

17.9 %

0.19 [ -0.12, 0.51 ]

10.9 %

0.26 [ -0.14, 0.67 ]

10.9 %

0.26 [ -0.14, 0.67 ]

11.0 %

-0.04 [ -0.44, 0.36 ]

11.0 %

-0.04 [ -0.44, 0.36 ]

11.0 %

0.11 [ -0.29, 0.51 ]

11.0 %

0.11 [ -0.29, 0.51 ]

2.2 %

0.63 [ -0.27, 1.53 ]

10

2.2 %

0.63 [ -0.27, 1.53 ]

390

100.0 %

0.15 [ 0.01, 0.28 ]

Heterogeneity: Chi2 = 0.13, df = 1 (P = 0.71); I2 =0.0% Test for overall effect: Z = 1.20 (P = 0.23) 7 Trail Making A Mäntynen 2013

Subtotal (95% CI)

58

-32.8 (11.6)

58

40

-36 (13)

40

Heterogeneity: not applicable Test for overall effect: Z = 1.26 (P = 0.21) 8 Trail Making B Mäntynen 2013

Subtotal (95% CI)

58

-80 (37.5)

58

40

-78.5 (37)

40

Heterogeneity: not applicable Test for overall effect: Z = 0.19 (P = 0.85) 9 Stroop/colour-word interference time Mäntynen 2013

Subtotal (95% CI)

58

-118.7 (33.2)

58

40

-122.5 (36.7)

40

Heterogeneity: not applicable Test for overall effect: Z = 0.53 (P = 0.60) 10 Test of Everyday Attention/incompatibility Brissart 2012

Subtotal (95% CI)

10

52.8 (4.49)

10

10

46.5 (12.82)


Total (95% CI)

504


-0.5

-0.25

Favours control


0

0.25

0.5


101

Analysis 3.2. Comparison 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control, Outcome 2 Information processing speed. Review:


Comparison: 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control Outcome: 2 Information processing speed

Study or subgroup

Intervention


Control

N

Mean(SD)

N

Mean(SD)

9

-44.4 (10.6)

Weight

IV,Fixed,95% CI


1 Sustained Attention Test/variation reaction time Tesar 2005

Subtotal (95% CI)

10

-44.4 (10.6)

10

12.3 %

0.0 [ -0.90, 0.90 ]

12.3 %

0.0 [ -0.90, 0.90 ]

20.5 %

0.25 [ -0.44, 0.95 ]

20.5 %

0.25 [ -0.44, 0.95 ]

20.7 %

0.07 [ -0.62, 0.77 ]

20.7 %

0.07 [ -0.62, 0.77 ]

20.5 %

0.25 [ -0.45, 0.94 ]

20.5 %

0.25 [ -0.45, 0.94 ]

26.0 %

-0.27 [ -0.89, 0.35 ]

25

26.0 %

-0.27 [ -0.89, 0.35 ]

121

100.0 %

0.05 [ -0.27, 0.36 ]

9

Heterogeneity: not applicable Test for overall effect: Z = 0.0 (P = 1.0) 2 2-back/reaction time Fink 2010

Subtotal (95% CI)

11

-589 (146)

11

29

-640.7 (215.2)

29

Heterogeneity: not applicable Test for overall effect: Z = 0.72 (P = 0.47) 3 Preference shifting/reaction time Fink 2010

Subtotal (95% CI)

11

-638 (185)

11

29

-649.2 (141.6)

29

Heterogeneity: not applicable Test for overall effect: Z = 0.20 (P = 0.84) 4 Response shifting/reaction time Fink 2010

Subtotal (95% CI)

11

-656 (219)

11

29

-702.4 (170.5)

29

Heterogeneity: not applicable Test for overall effect: Z = 0.69 (P = 0.49) 5 Test Battery of Attention/object alternation Hildebrandt 2007

17

Subtotal (95% CI)

17

-820 (323)

25

-744 (233)


Total (95% CI)

60


-1

-0.5

Favours control


0

0.5

1


102

Analysis 3.3. Comparison 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control, Outcome 3 Memory span. Review:


Comparison: 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control Outcome: 3 Memory span

Study or subgroup

Intervention

Mean Difference

Control

N

Mean(SD)

N

Mean(SD)

Brissart 2012

10

5.9 (0.74)

10

5.5 (1.18)

Total (95% CI)

10

Weight

IV,Fixed,95% CI


1 WAIS/digit span forward

10

100.0 %

0.40 [ -0.46, 1.26 ]

100.0 %

0.40 [ -0.46, 1.26 ]


-2

-1

Favours control


0

1

2


103

Analysis 3.4. Comparison 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control, Outcome 4 Working memory. Review:


Comparison: 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control Outcome: 4 Working memory

Study or subgroup

Intervention


Control

N

Mean(SD)

N

Mean(SD)

11

4.2 (6.5)

29

3 (4.2)

Weight

IV,Fixed,95% CI


1 2-back/commissions Fink 2010

Subtotal (95% CI)

11

45.3 %

0.24 [ -0.46, 0.94 ]

45.3 %

0.24 [ -0.46, 0.94 ]

28.4 %

-0.20 [ -1.08, 0.68 ]

28.4 %

-0.20 [ -1.08, 0.68 ]

26.3 %

0.75 [ -0.16, 1.67 ]

10

26.3 %

0.75 [ -0.16, 1.67 ]

49

100.0 %

0.25 [ -0.22, 0.72 ]

29

Heterogeneity: not applicable Test for overall effect: Z = 0.67 (P = 0.50) 2 WAIS/digit span backward Brissart 2012

Subtotal (95% CI)

10

4.3 (1.16)

10

10

4.6 (1.65)

10

Heterogeneity: not applicable Test for overall effect: Z = 0.45 (P = 0.65) 3 Test of Everyday Attention/working memory Brissart 2012

Subtotal (95% CI)

10

12.1 (1.73)

10

10

9.8 (3.77)


Total (95% CI)

31


-2

-1

Favours control


0

1

2


104

Analysis 3.5. Comparison 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control, Outcome 5 Immediate verbal memory. Review:


Comparison: 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control Outcome: 5 Immediate verbal memory

Study or subgroup

Intervention


Control

N

Mean(SD)

N

Mean(SD)

10

52 (8.2)

9

48.2 (13.1)

Weight

IV,Fixed,95% CI


1 Verbal Learning Test Tesar 2005

Subtotal (95% CI)

10

9

6.5 %

0.34 [ -0.57, 1.25 ]

6.5 %

0.34 [ -0.57, 1.25 ]

Heterogeneity: not applicable Test for overall effect: Z = 0.73 (P = 0.47) 2 California Verbal Learning Test/learning trials Fink 2010

11

12.1 (2.1)

29

11.5 (1.7)

11.0 %

0.32 [ -0.37, 1.02 ]

Hildebrandt 2007

17

12.29 (2.12)

25

11.3 (1.94)

13.6 %

0.48 [ -0.14, 1.11 ]

Subtotal (95% CI)

28

24.6 %

0.41 [ -0.05, 0.88 ]

20.8 %

0.16 [ -0.34, 0.67 ]

20.8 %

0.16 [ -0.34, 0.67 ]

9.7 %

0.08 [ -0.66, 0.82 ]

9.7 %

0.08 [ -0.66, 0.82 ]

32.7 %

0.24 [ -0.16, 0.65 ]

32.7 %

0.24 [ -0.16, 0.65 ]

5.7 %

1.17 [ 0.21, 2.14 ]

5.7 %

1.17 [ 0.21, 2.14 ]

54

Heterogeneity: Chi2 = 0.11, df = 1 (P = 0.74); I2 =0.0% Test for overall effect: Z = 1.73 (P = 0.083) 3 California Verbal Learning Test/total Stuifbergen 2012

Subtotal (95% CI)

34

52.2 (12.3)

34

27

50.2 (12.1)

27

Heterogeneity: not applicable Test for overall effect: Z = 0.63 (P = 0.53) 4 Hopkins Verbal Learning Test - revised Chiaravalloti 2005

14

Subtotal (95% CI)

14

26.57 (3.69)

14

26.29 (2.89)

14

Heterogeneity: not applicable Test for overall effect: Z = 0.22 (P = 0.83) 5 Buschke Selective Reminding Test/consistent long-term retrieval Mäntynen 2013

Subtotal (95% CI)

58

45.1 (16.8)

58

40

41.1 (15.8)

40

Heterogeneity: not applicable Test for overall effect: Z = 1.17 (P = 0.24) 6 Buschke Selective Reminding Test/free recall Brissart 2012

Subtotal (95% CI)

10

10

11.61 (1.12)

10

9.37 (2.33)

10


-1

-0.5

Favours control

0

0.5

1


(Continued . . . )


105


Intervention N

Total (95% CI)


Control Mean(SD)

154

N

Mean(SD)

Weight

IV,Fixed,95% CI


IV,Fixed,95% CI

154

100.0 %

0.31 [ 0.08, 0.54 ]


-1

-0.5

0

Favours control

0.5

1


Analysis 3.6. Comparison 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control, Outcome 6 Immediate visual memory. Review:


Comparison: 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control Outcome: 6 Immediate visual memory

Study or subgroup

Intervention


Control

N

Mean(SD)

N

Mean(SD)

10

52 (8.2)

9

48.2 (13.1)

Weight

IV,Random,95% CI


1 Non-Verbal Learning Test Tesar 2005

Subtotal (95% CI)

10

9

18.1 %

0.34 [ -0.57, 1.25 ]

18.1 %

0.34 [ -0.57, 1.25 ]

31.1 %

-0.04 [ -0.54, 0.47 ]

31.1 %

-0.04 [ -0.54, 0.47 ]

Heterogeneity: not applicable Test for overall effect: Z = 0.73 (P = 0.47) 2 Brief Visuospatial Memory Test - revised/total Stuifbergen 2012

Subtotal (95% CI)

34

23.8 (7.6)

34

27

24.1 (7.8)

27

Heterogeneity: not applicable Test for overall effect: Z = 0.15 (P = 0.88) 3 10/36 Spatial Recall/total correct Brissart 2012

10

20.33 (3.32)

10

14.3 (4.27)

15.7 %

1.51 [ 0.49, 2.53 ]

Mäntynen 2013

58

22.4 (4.9)

40

21.4 (4.4)

35.1 %

0.21 [ -0.19, 0.61 ]

50.8 %

0.77 [ -0.49, 2.03 ]

Subtotal (95% CI)

68

50 -1

-0.5

Favours control

0

0.5

1


(Continued . . . )


106


Intervention N

Heterogeneity:

Tau2

= 0.69;

Chi2


Control Mean(SD)

= 5.38, df = 1 (P = 0.02);

N I2

Mean(SD)

Weight

IV,Random,95% CI


IV,Random,95% CI

=81%

Test for overall effect: Z = 1.20 (P = 0.23)

Total (95% CI)

112

86

100.0 %

0.36 [ -0.14, 0.86 ]

Heterogeneity: Tau2 = 0.14; Chi2 = 7.18, df = 3 (P = 0.07); I2 =58% Test for overall effect: Z = 1.42 (P = 0.16) Test for subgroup differences: Chi2 = 1.62, df = 2 (P = 0.44), I2 =0.0%

-1

-0.5

0

Favours control

0.5

1


Analysis 3.7. Comparison 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control, Outcome 7 Delayed memory. Review:


Comparison: 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control Outcome: 7 Delayed memory

Study or subgroup

Intervention N


Control Mean(SD)

N

Mean(SD)

Weight

IV,Fixed,95% CI


1 California Verbal Learning Test/long delay free recall Hildebrandt 2007

17

13.24 (3.35)

25

12.16 (3.22)

10.5 %

0.32 [ -0.30, 0.94 ]

Stuifbergen 2012

34

12.3 (3.6)

27

10.7 (4.1)

15.4 %

0.41 [ -0.10, 0.92 ]

25.9 %

0.38 [ -0.02, 0.77 ]

15.8 %

0.06 [ -0.44, 0.57 ]

15.8 %

0.06 [ -0.44, 0.57 ]

Subtotal (95% CI)

51

52

Heterogeneity: Chi2 = 0.05, df = 1 (P = 0.83); I2 =0.0% Test for overall effect: Z = 1.87 (P = 0.061) 2 Brief Visuospatial Memory Test - revised/delayed Stuifbergen 2012

Subtotal (95% CI)

34

34

9.3 (3)

27

9.1 (3.1)

27

Heterogeneity: not applicable Test for overall effect: Z = 0.25 (P = 0.80) 3 Buschke Selective Reminding Test/delayed recall

-2

-1

Favours control

0

1

2


(Continued . . . )


107


Intervention


Control

Weight

N

Mean(SD)

N

Mean(SD)

Brissart 2012

10

13.4 (1.65)

10

10.7 (3.33)

4.6 %

0.98 [ 0.04, 1.92 ]

Mäntynen 2013

58

10 (2.1)

40

10 (1.7)

24.8 %

0.0 [ -0.40, 0.40 ]

29.4 %

0.15 [ -0.22, 0.52 ]

Subtotal (95% CI)

68

IV,Fixed,95% CI


IV,Fixed,95% CI

50

Heterogeneity: Chi2 = 3.55, df = 1 (P = 0.06); I2 =72% Test for overall effect: Z = 0.81 (P = 0.42) 4 10/36 Spatial Recall/delayed Brissart 2012

10

7.7 (1.77)

10

4.8 (2.62)

4.2 %

1.24 [ 0.27, 2.22 ]

Mäntynen 2013

58

7.8 (2.2)

40

7.7 (1.8)

24.8 %

0.05 [ -0.35, 0.45 ]

50

29.0 %

0.22 [ -0.15, 0.59 ]

179

100.0 %

0.22 [ 0.02, 0.42 ]

Subtotal (95% CI)

68


Total (95% CI)

221


-2

-1

Favours control


0

1

2


108

Analysis 3.8. Comparison 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control, Outcome 8 Executive functions. Review:


Comparison: 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control Outcome: 8 Executive functions

Study or subgroup

Intervention N


Control Mean(SD)

N

Mean(SD)

9

47.5 (18.6)

Weight

IV,Fixed,95% CI


1 Computer-aided Card-Sorting Procedure/correct answers Tesar 2005

Subtotal (95% CI)

10

47.1 (19.9)

10

11.8 %

-0.02 [ -0.92, 0.88 ]

11.8 %

-0.02 [ -0.92, 0.88 ]

19.6 %

-0.33 [ -1.03, 0.37 ]

19.6 %

-0.33 [ -1.03, 0.37 ]

19.8 %

0.18 [ -0.52, 0.87 ]

19.8 %

0.18 [ -0.52, 0.87 ]

37.0 %

-0.30 [ -0.81, 0.21 ]

37.0 %

-0.30 [ -0.81, 0.21 ]

11.8 %

0.57 [ -0.32, 1.47 ]

10

11.8 %

0.57 [ -0.32, 1.47 ]

104

100.0 %

-0.08 [ -0.39, 0.23 ]

9

Heterogeneity: not applicable Test for overall effect: Z = 0.04 (P = 0.97) 2 Preference shifting/trials to criterion Fink 2010

Subtotal (95% CI)

11

33 (19)

11

29

39.8 (20.5)

29

Heterogeneity: not applicable Test for overall effect: Z = 0.93 (P = 0.35) 3 Response shifting/trials to criterion Fink 2010

Subtotal (95% CI)

11

49.3 (23.7)

11

29

44.7 (26.1)

29

Heterogeneity: not applicable Test for overall effect: Z = 0.50 (P = 0.62) 4 Sorting Test from the Delis-Kaplan Executive Function System Stuifbergen 2012

Subtotal (95% CI)

34

9.2 (2.7)

34

27

10.1 (3.2)

27

Heterogeneity: not applicable Test for overall effect: Z = 1.17 (P = 0.24) 5 Test of Everyday Attention/flexibility errors Brissart 2012

Subtotal (95% CI)

10

-3.9 (5.2)

10

10

-8.1 (8.44)


Total (95% CI)

76


-1

-0.5

Favours control


0

0.5

1


109

Analysis 3.9. Comparison 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control, Outcome 9 Visual functions. Review:


Comparison: 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control Outcome: 9 Visual functions

Study or subgroup

Intervention


Control

N

Mean(SD)

N

Mean(SD)

10

11.2 (2.9)

9

9 (2.4)

Weight

IV,Fixed,95% CI


1 Intelligence Test-revised (HAWIE-R) Tesar 2005

Subtotal (95% CI)

10

22.5 %

0.79 [ -0.16, 1.73 ]

22.5 %

0.79 [ -0.16, 1.73 ]

77.5 %

0.29 [ -0.21, 0.80 ]

27

77.5 %

0.29 [ -0.21, 0.80 ]

36

100.0 %

0.40 [ -0.04, 0.85 ]

9

Heterogeneity: not applicable Test for overall effect: Z = 1.63 (P = 0.10) 2 Judgement of Line Orientation Test Stuifbergen 2012

Subtotal (95% CI)

34

27.5 (3.6)

34

27

26.2 (5.2)


Total (95% CI)

44


-1

-0.5

Favours control


0

0.5

1


110

Analysis 3.10. Comparison 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control, Outcome 10 Verbal functions. Review:


Comparison: 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control Outcome: 10 Verbal functions

Study or subgroup

Intervention N


Control Mean(SD)

N

Mean(SD)

Weight

IV,Fixed,95% CI


1 Controlled Oral Word Association Test (semantic fluency) Brissart 2012

10

18.7 (2.83)

10

15.7 (4.76)

8.8 %

0.73 [ -0.18, 1.65 ]

Mäntynen 2013

58

24.5 (6.3)

40

22.8 (7.9)

44.6 %

0.24 [ -0.16, 0.65 ]

Stuifbergen 2012

34

35.8 (10.6)

27

35.3 (11.7)

28.6 %

0.04 [ -0.46, 0.55 ]

81.9 %

0.23 [ -0.07, 0.52 ]

9.5 %

0.05 [ -0.83, 0.93 ]

9.5 %

0.05 [ -0.83, 0.93 ]

8.6 %

0.81 [ -0.11, 1.73 ]

10

8.6 %

0.81 [ -0.11, 1.73 ]

97

100.0 %

0.26 [ -0.01, 0.53 ]

Subtotal (95% CI)

102

77

Heterogeneity: Chi2 = 1.69, df = 2 (P = 0.43); I2 =0.0% Test for overall effect: Z = 1.48 (P = 0.14) 2 Controlled Oral Word Association Test (phonologic fluency) Brissart 2012

Subtotal (95% CI)

10

10.7 (4.37)

10

10

10.5 (3.06)

10

Heterogeneity: not applicable Test for overall effect: Z = 0.11 (P = 0.91) 3 Boston Naming Test Brissart 2012

Subtotal (95% CI)

10

50.9 (4.61)

10

10

46.9 (4.82)


Total (95% CI)

122


-1

-0.5

Favours control


0

0.5

1


111

Analysis 3.11. Comparison 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control, Outcome 11 Everyday cognitive performance/patient’s report. Review:


Comparison: 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control Outcome: 11 Everyday cognitive performance/patient’s report

Study or subgroup

Intervention


Control

N

Mean(SD)

N

Mean(SD)

72

10.3 (6.6)

153

10.8 (6.4)

Weight

IV,Fixed,95% CI


1 Memory Aids Questionnaire Lincoln 2002

Subtotal (95% CI)

72

153

21.6 %

-0.08 [ -0.36, 0.20 ]

21.6 %

-0.08 [ -0.36, 0.20 ]

3.0 %

-0.40 [ -1.15, 0.35 ]

3.0 %

-0.40 [ -1.15, 0.35 ]

21.0 %

-0.02 [ -0.31, 0.26 ]

21.0 %

-0.02 [ -0.31, 0.26 ]

21.3 %

0.01 [ -0.27, 0.30 ]

21.3 %

0.01 [ -0.27, 0.30 ]

6.6 %

0.30 [ -0.21, 0.81 ]

6.6 %

0.30 [ -0.21, 0.81 ]

Heterogeneity: not applicable Test for overall effect: Z = 0.54 (P = 0.59) 2 Memory Functioning Questionnaire/overall rating Chiaravalloti 2005

14

Subtotal (95% CI)

14

3.46 (1.56)

14

4.14 (1.75)

14

Heterogeneity: not applicable Test for overall effect: Z = 1.04 (P = 0.30) 3 Everyday Memory Questionnaire/patient’s report Lincoln 2002

Subtotal (95% CI)

71

23.3 (20.3)

71

146

23.8 (21.4)

146

Heterogeneity: not applicable Test for overall effect: Z = 0.16 (P = 0.87) 4 Dysexecutive Memory Questionnaire/patient’s report Lincoln 2002

Subtotal (95% CI)

71

20.4 (13)

71

152

20.2 (14.4)

152

Heterogeneity: not applicable Test for overall effect: Z = 0.10 (P = 0.92) 5 Strategy Subscale of the Multifactorial Memory Questionnaire Stuifbergen 2012

Subtotal (95% CI)

34

43.63 (11.32)

34

27

40.43 (9.34)

27

Heterogeneity: not applicable Test for overall effect: Z = 1.16 (P = 0.25) 6 Multiple Sclerosis Neuropsychological Questionnaire Mäntynen 2013

58

-24.2 (8.2)

40

-29.1 (10.2)

10.1 %

0.54 [ 0.13, 0.95 ]

Stuifbergen 2012

34

-29.68 (10.74)

27

-27.92 (11.11)

6.6 %

-0.16 [ -0.67, 0.35 ]

16.7 %

0.26 [ -0.06, 0.58 ]

Subtotal (95% CI)

92

67


-1

-0.5

Favours control

0

0.5

1


(Continued . . . )


112


Intervention


Control

N

Mean(SD)

N

Mean(SD)

58

-28.7 (12)

40

-37.3 (13)

Weight

IV,Fixed,95% CI


IV,Fixed,95% CI

7 Perceived Deficits Questionnaire Mäntynen 2013

Subtotal (95% CI)

58

9.8 %

0.69 [ 0.27, 1.10 ]

40

9.8 %

0.69 [ 0.27, 1.10 ]

599

100.0 %

0.10 [ -0.03, 0.23 ]


Total (95% CI)

412


-1

-0.5

Favours control


0

0.5

1


113

Analysis 3.12. Comparison 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control, Outcome 12 Everyday cognitive performance/carer’s report. Review:


Comparison: 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control Outcome: 12 Everyday cognitive performance/carer’s report

Study or subgroup

Intervention N


Control Mean(SD)

N

Mean(SD)

122

20.5 (22.6)

Weight

IV,Fixed,95% CI


1 Everyday Memory Questionnaire/carer’s report Lincoln 2002

Subtotal (95% CI)

55

21.4 (21.4)

55

38.2 %

0.04 [ -0.28, 0.36 ]

38.2 %

0.04 [ -0.28, 0.36 ]

38.0 %

-0.01 [ -0.33, 0.31 ]

38.0 %

-0.01 [ -0.33, 0.31 ]

23.8 %

0.08 [ -0.32, 0.49 ]

40

23.8 %

0.08 [ -0.32, 0.49 ]

287

100.0 %

0.03 [ -0.17, 0.23 ]

122

Heterogeneity: not applicable Test for overall effect: Z = 0.25 (P = 0.80) 2 Dysexecutive Syndrome Questionnaire/carer’s report Lincoln 2002

Subtotal (95% CI)

54

21.4 (16.7)

54

125

21.6 (16.9)

125

Heterogeneity: not applicable Test for overall effect: Z = 0.07 (P = 0.94) 3 Multiple Sclerosis Neuropsychological Questionnaire - Informant Mäntynen 2013

Subtotal (95% CI)

58

-20.2 (10.4)

58

40

-21.1 (11.3)


Total (95% CI)

167


-0.5

-0.25

Favours control


0

0.25

0.5


114

Analysis 3.13. Comparison 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control, Outcome 13 Depression. Review:


Comparison: 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control Outcome: 13 Depression

Study or subgroup

Intervention

Mean Difference

Control

Weight

IV,Fixed,95% CI

Mean Difference

N

Mean(SD)

N

Mean(SD)

IV,Fixed,95% CI

Chiaravalloti 2005

14

-6.21 (6.2)

14

-8.36 (6.28)

26.0 %

2.15 [ -2.47, 6.77 ]

Hildebrandt 2007

17

-10.3 (8.5)

25

-11 (7.9)

21.4 %

0.70 [ -4.39, 5.79 ]

Mäntynen 2013

58

-10.2 (7)

40

-10.4 (67)

1.3 %

0.20 [ -20.64, 21.04 ]

Tesar 2005

10

-8.6 (4.1)

9

-7.7 (3.2)

51.3 %

-0.90 [ -4.19, 2.39 ]

100.0 %

0.25 [ -2.11, 2.61 ]

1 Beck Depression Inventory

Total (95% CI)

99

88

Heterogeneity: Chi2 = 1.15, df = 3 (P = 0.77); I2 =0.0% Test for overall effect: Z = 0.21 (P = 0.84) Test for subgroup differences: Not applicable

-20

-10

Favours control


0

10

20


115

Analysis 3.14. Comparison 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control, Outcome 14 Quality of life. Review:


Comparison: 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control Outcome: 14 Quality of life

Study or subgroup

Intervention N


Control Mean(SD)

N

Mean(SD)

25

47.8 (9.7)

Weight

IV,Fixed,95% CI


1 SF-36 Health Questionnaire (short form)/mental score Hildebrandt 2007

17

Subtotal (95% CI)

17

48.5 (13.3)

30.0 %

0.06 [ -0.56, 0.68 ]

30.0 %

0.06 [ -0.56, 0.68 ]

70.0 %

0.20 [ -0.20, 0.61 ]

40

70.0 %

0.20 [ -0.20, 0.61 ]

65

100.0 %

0.16 [ -0.18, 0.50 ]

25

Heterogeneity: not applicable Test for overall effect: Z = 0.19 (P = 0.85) 2 Brief version of the World Health Organization Quality of Life/psychological total score Mäntynen 2013

Subtotal (95% CI)

58

14.3 (2.6)

58

40

13.8 (2.2)


Total (95% CI)

75


-2

-1

Favours control


0

1

2


116

Analysis 3.15. Comparison 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control, Outcome 15 Fatigue. Review:


Comparison: 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control Outcome: 15 Fatigue

Study or subgroup

Intervention


Control

N

Mean(SD)

N

Mean(SD)

10

-41.2 (13.2)

9

-33.3 (16.5)

Weight

IV,Random,95% CI


1 Modified Fatigue Impact Scale Tesar 2005

Subtotal (95% CI)

10

20.9 %

-0.51 [ -1.43, 0.41 ]

20.9 %

-0.51 [ -1.43, 0.41 ]

33.3 %

-0.05 [ -0.66, 0.57 ]

33.3 %

-0.05 [ -0.66, 0.57 ]

45.7 %

0.43 [ 0.03, 0.84 ]

40

45.7 %

0.43 [ 0.03, 0.84 ]

74

100.0 %

0.08 [ -0.44, 0.59 ]

9

Heterogeneity: not applicable Test for overall effect: Z = 1.08 (P = 0.28) 2 Fatigue Severity Scale Hildebrandt 2007

17

Subtotal (95% CI)

17

-37.5 (15)

25

-36.8 (14.5)

25

Heterogeneity: not applicable Test for overall effect: Z = 0.15 (P = 0.88) 3 Fatigue Scale for Motor and Cognitive Fatigue/total Mäntynen 2013

Subtotal (95% CI)

58

58

-61.8 (17.7)

40

-69 (14.5)


Total (95% CI)

85


-2

-1

Favours control


0

1

2


117

Analysis 3.16. Comparison 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control, Outcome 16 Anxiety. Review:


Comparison: 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control Outcome: 16 Anxiety

Study or subgroup

Intervention

Mean Difference

Control

N

Mean(SD)

N

Mean(SD)

-30.14 (10.04)

14

-32.5 (9.31)

Weight

IV,Fixed,95% CI


1 State-Trait Anxiety Inventory/state Chiaravalloti 2005

14

Total (95% CI)

14

14

100.0 %

2.36 [ -4.81, 9.53 ]

100.0 %

2.36 [ -4.81, 9.53 ]


-4

-2

Favours control


0

2

4


118

Analysis 3.17. Comparison 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control, Outcome 17 Impact of the disease. Review:


Comparison: 3 Cognitive training combined with other neuropsychological rehabilitation methods versus any control Outcome: 17 Impact of the disease

Study or subgroup

Intervention N


Control Mean(SD)

N

Mean(SD)

40

-26.8 (15.8)

Weight

IV,Fixed,95% CI


1 Multiple Sclerosis Impact Scale/psychological total score Mäntynen 2013

Subtotal (95% CI)

58

-26.4 (15.4)

58

61.1 %

40

0.03 [ -0.38, 0.43 ]

61.1 % 0.03 [ -0.38, 0.43 ]

Heterogeneity: not applicable Test for overall effect: Z = 0.12 (P = 0.90) 2 Control subscale of the Multiple Sclerosis Self Efficacy Scale Stuifbergen 2012

Subtotal (95% CI)

34

553.24 (167.58)

34

27

38.9 %

540.19 (203.25)

0.07 [ -0.44, 0.58 ]

27

38.9 % 0.07 [ -0.44, 0.58 ]

67

100.0 % 0.04 [ -0.27, 0.36 ]


Total (95% CI)

92


-1

-0.5

Favours control


0

0.5

1


119

Analysis 4.1. Comparison 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up), Outcome 1 Attention. Review:


Comparison: 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up) Outcome: 1 Attention

Study or subgroup

Intervention N


Control Mean(SD)

N

Mean(SD)

42.1 (12.6)

9

53.9 (21.5)

Weight

IV,Fixed,95% CI


1 Sustained attention test/correct answers Tesar 2005

Subtotal (95% CI)

10

10

9

2.3 %

-0.65 [ -1.58, 0.28 ]

2.3 %

-0.65 [ -1.58, 0.28 ]

Heterogeneity: not applicable Test for overall effect: Z = 1.37 (P = 0.17) 2 Paced Auditory Serial Addition Test 3’ Mäntynen 2013

58

46.7 (11.8)

40

43.5 (11)

12.3 %

0.28 [ -0.13, 0.68 ]

Stuifbergen 2012

34

47.4 (9.6)

27

47.2 (10.7)

7.9 %

0.02 [ -0.49, 0.52 ]

20.2 %

0.18 [ -0.14, 0.49 ]

Subtotal (95% CI)

92

67

Heterogeneity: Chi2 = 0.61, df = 1 (P = 0.44); I2 =0.0% Test for overall effect: Z = 1.09 (P = 0.27) 3 Paced Auditory Serial Addition Test 2’ Mäntynen 2013

58

32.9 (12.1)

40

30.8 (10.3)

12.4 %

0.18 [ -0.22, 0.59 ]

Stuifbergen 2012

34

34.2 (9.8)

27

38.1 (9.8)

7.7 %

-0.39 [ -0.90, 0.12 ]

20.1 %

-0.04 [ -0.36, 0.28 ]

Subtotal (95% CI)

92

67

Heterogeneity: Chi2 = 3.01, df = 1 (P = 0.08); I2 =67% Test for overall effect: Z = 0.24 (P = 0.81) 4 Symbol Digit Modalities Test Mäntynen 2013

58

50.6 (12.1)

40

48.2 (8.2)

12.3 %

0.22 [ -0.18, 0.63 ]

Stuifbergen 2012

34

49.7 (12.7)

27

50.6 (13.1)

7.9 %

-0.07 [ -0.57, 0.44 ]

20.2 %

0.11 [ -0.21, 0.42 ]

12.4 %

-0.10 [ -0.50, 0.31 ]

12.4 %

-0.10 [ -0.50, 0.31 ]

12.4 %

-0.10 [ -0.50, 0.30 ]

Subtotal (95% CI)

92

67

Heterogeneity: Chi2 = 0.78, df = 1 (P = 0.38); I2 =0.0% Test for overall effect: Z = 0.68 (P = 0.50) 5 Trail Making A Mäntynen 2013

Subtotal (95% CI)

58

-32.1 (12.4)

58

40

-31 (9.2)

40

Heterogeneity: not applicable Test for overall effect: Z = 0.47 (P = 0.64) 6 Trail Making B Mäntynen 2013

58

-79.1 (36.4)

40

-75.4 (35.6) -1

-0.5

Favours control

0

0.5

1


(Continued . . . )


120


Intervention N

Subtotal (95% CI)


Control Mean(SD)

58

N

Mean(SD)

Weight

IV,Fixed,95% CI


IV,Fixed,95% CI

40

12.4 %

-0.10 [ -0.50, 0.30 ]

12.4 %

-0.01 [ -0.41, 0.40 ]

40

12.4 %

-0.01 [ -0.41, 0.40 ]

330

100.0 %

0.01 [ -0.13, 0.15 ]

Heterogeneity: not applicable Test for overall effect: Z = 0.49 (P = 0.62) 7 Stroop/colour-word interference time Mäntynen 2013

Subtotal (95% CI)

58

-116.2 (36.2)

58

40

-116 (30.3)


Total (95% CI)

460


-1

-0.5

Favours control


0

0.5

1


121

Analysis 4.2. Comparison 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up), Outcome 2 Information processing speed. Review:


Comparison: 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up) Outcome: 2 Information processing speed

Study or subgroup

Intervention N


Control Mean(SD)

N

Mean(SD)

9

-50.7 (7.7)

Weight

IV,Fixed,95% CI


1 Sustained attention test/variation reaction time Tesar 2005

Subtotal (95% CI)

10

-44.8 (11.7)

10

26.6 %

0.56 [ -0.36, 1.49 ]

26.6 %

0.56 [ -0.36, 1.49 ]

23.9 %

-0.51 [ -1.49, 0.46 ]

23.9 %

-0.51 [ -1.49, 0.46 ]

24.8 %

-0.04 [ -1.00, 0.91 ]

24.8 %

-0.04 [ -1.00, 0.91 ]

24.7 %

0.14 [ -0.82, 1.09 ]

14

24.7 %

0.14 [ -0.82, 1.09 ]

51

100.0 %

0.05 [ -0.43, 0.53 ]

9

Heterogeneity: not applicable Test for overall effect: Z = 1.19 (P = 0.23) 2 2-back/reaction time Fink 2010

Subtotal (95% CI)

6

-685 (184)

6

14

-584.9 (187.5)

14

Heterogeneity: not applicable Test for overall effect: Z = 1.04 (P = 0.30) 3 Preference shifting/reaction time Fink 2010

Subtotal (95% CI)

6

-685 (142)

6

14

-676.6 (203.5)

14

Heterogeneity: not applicable Test for overall effect: Z = 0.09 (P = 0.93) 4 Response shifting/reaction time Fink 2010

Subtotal (95% CI)

6

-684 (230)

6

14

-716.1 (226)


Total (95% CI)

28


-2

-1

Favours control


0

1

2


122

Analysis 4.3. Comparison 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up), Outcome 3 Working memory. Review:


Comparison: 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up) Outcome: 3 Working memory

Study or subgroup

Intervention

Mean Difference

Control

N

Mean(SD)

N

Mean(SD)

6

4.2 (5.2)

14

3.1 (2.2)

Weight

IV,Fixed,95% CI


1 2-back/commissions Fink 2010

Total (95% CI)

6

14

100.0 %

1.10 [ -3.22, 5.42 ]

100.0 %

1.10 [ -3.22, 5.42 ]


-4

-2

Favours control


0

2

4


123

Analysis 4.4. Comparison 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up), Outcome 4 Immediate verbal memory. Review:


Comparison: 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up) Outcome: 4 Immediate verbal memory

Study or subgroup

Intervention


Control

N

Mean(SD)

N

Mean(SD)

10

56.9 (13.1)

9

50.4 (13.6)

Weight

IV,Fixed,95% CI


1 Verbal Learning Test Tesar 2005

Subtotal (95% CI)

8.5 %

0.47 [ -0.45, 1.38 ]

8.5 %

0.47 [ -0.45, 1.38 ]

12.9 %

-0.12 [ -0.87, 0.62 ]

12.9 %

-0.12 [ -0.87, 0.62 ]

7.6 %

0.41 [ -0.55, 1.38 ]

7.6 %

0.41 [ -0.55, 1.38 ]

27.5 %

0.33 [ -0.18, 0.84 ]

27.5 %

0.33 [ -0.18, 0.84 ]

43.5 %

0.26 [ -0.14, 0.67 ]

40

43.5 %

0.26 [ -0.14, 0.67 ]

104

100.0 %

0.26 [ -0.01, 0.53 ]

10

9

Heterogeneity: not applicable Test for overall effect: Z = 1.00 (P = 0.32) 2 Hopkins Verbal Learning Test - revised Chiaravalloti 2005

14

Subtotal (95% CI)

14

27.07 (5.15)

14

27.64 (3.61)

14

Heterogeneity: not applicable Test for overall effect: Z = 0.33 (P = 0.74) 3 California Verbal Learning Test/learning trials Fink 2010

6

Subtotal (95% CI)

12.5 (2.1)

14

6

11.8 (1.4)

14

Heterogeneity: not applicable Test for overall effect: Z = 0.84 (P = 0.40) 4 California Verbal Learning Test/total Stuifbergen 2012

Subtotal (95% CI)

34

58.4 (13.6)

27

34

53.8 (14.3)

27

Heterogeneity: not applicable Test for overall effect: Z = 1.26 (P = 0.21) 5 Buschke Selective Reminding Test/consistent long-term retrieval Mäntynen 2013

Subtotal (95% CI)

58

50.2 (18.2)

58

40

45.7 (15.2)


Total (95% CI)

122


-2

-1

Favours control


0

1

2


124

Analysis 4.5. Comparison 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up), Outcome 5 Immediate visual memory. Review:


Comparison: 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up) Outcome: 5 Immediate visual memory

Study or subgroup

Intervention


Control

N

Mean(SD)

N

Mean(SD)

10

49 (14.9)

9

48.3 (12.2)

Weight

IV,Fixed,95% CI


1 Non-Verbal Learning Test Tesar 2005

Subtotal (95% CI)

10

11.2 %

0.05 [ -0.85, 0.95 ]

11.2 %

0.05 [ -0.85, 0.95 ]

35.5 %

0.05 [ -0.46, 0.55 ]

35.5 %

0.05 [ -0.46, 0.55 ]

53.3 %

0.62 [ 0.21, 1.03 ]

40

53.3 %

0.62 [ 0.21, 1.03 ]

76

100.0 %

0.35 [ 0.05, 0.65 ]

9

Heterogeneity: not applicable Test for overall effect: Z = 0.11 (P = 0.92) 2 Brief Visuospatial Memory Test - revised/total Stuifbergen 2012

Subtotal (95% CI)

34

24.9 (6)

34

27

24.6 (6.9)

27

Heterogeneity: not applicable Test for overall effect: Z = 0.18 (P = 0.86) 3 10/36 Spatial Recall/total correct Mäntynen 2013

Subtotal (95% CI)

58

23.8 (4.5)

58

40

20.9 (4.8)


Total (95% CI)

102


-2

-1

Favours control


0

1

2


125

Analysis 4.6. Comparison 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up), Outcome 6 Delayed memory. Review:


Comparison: 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up) Outcome: 6 Delayed memory

Study or subgroup

Intervention


Control

N

Mean(SD)

N

Mean(SD)

34

12.5 (4.1)

27

11.4 (4.1)

Weight

IV,Fixed,95% CI


1 California Verbal Learning Test/delay Stuifbergen 2012

Subtotal (95% CI)

34

19.6 %

0.26 [ -0.24, 0.77 ]

19.6 %

0.26 [ -0.24, 0.77 ]

19.7 %

0.20 [ -0.30, 0.71 ]

19.7 %

0.20 [ -0.30, 0.71 ]

30.9 %

0.20 [ -0.21, 0.60 ]

30.9 %

0.20 [ -0.21, 0.60 ]

29.8 %

0.57 [ 0.16, 0.99 ]

40

29.8 %

0.57 [ 0.16, 0.99 ]

134

100.0 %

0.32 [ 0.10, 0.55 ]

27

Heterogeneity: not applicable Test for overall effect: Z = 1.02 (P = 0.31) 2 Brief Visuospatial Memory Test - revised/delayed Stuifbergen 2012

Subtotal (95% CI)

34

9.3 (2.1)

34

27

8.8 (2.8)

27

Heterogeneity: not applicable Test for overall effect: Z = 0.78 (P = 0.43) 3 Buschke Selective Reminding Test/delayed Mäntynen 2013

Subtotal (95% CI)

58

10.4 (2.2)

58

40

10 (1.7)

40

Heterogeneity: not applicable Test for overall effect: Z = 0.96 (P = 0.34) 4 10/36 Spatial Recall/delayed Mäntynen 2013

Subtotal (95% CI)

58

8.5 (1.9)

58

40

7.4 (1.9)


Total (95% CI)

184


-1

-0.5

Favours control


0

0.5

1


126

Analysis 4.7. Comparison 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up), Outcome 7 Executive functions. Review:


Comparison: 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up) Outcome: 7 Executive functions

Study or subgroup

Intervention N


Control Mean(SD)

N

Mean(SD)

9

53.9 (21.5)

Weight

IV,Random,95% CI


1 Computer-aided Card-Sorting Procedure/correct answers Tesar 2005

Subtotal (95% CI)

10

42.1 (12.6)

10

20.7 %

-0.65 [ -1.58, 0.28 ]

20.7 %

-0.65 [ -1.58, 0.28 ]

18.9 %

0.75 [ -0.24, 1.74 ]

18.9 %

0.75 [ -0.24, 1.74 ]

19.6 %

-0.37 [ -1.34, 0.59 ]

19.6 %

-0.37 [ -1.34, 0.59 ]

40.9 %

-0.29 [ -0.80, 0.22 ]

27

40.9 %

-0.29 [ -0.80, 0.22 ]

64

100.0 %

-0.18 [ -0.69, 0.32 ]

9

Heterogeneity: not applicable Test for overall effect: Z = 1.37 (P = 0.17) 2 Preference shifting/trials to criterion Fink 2010

Subtotal (95% CI)

6

59.2 (22.5)

6

14

42.3 (21.4)

14

Heterogeneity: not applicable Test for overall effect: Z = 1.48 (P = 0.14) 3 Response shifting/trials to criterion Fink 2010

Subtotal (95% CI)

6

40.4 (31.6)

6

14

50.8 (24.6)

14

Heterogeneity: not applicable Test for overall effect: Z = 0.76 (P = 0.45) 4 Sorting Test from the Delis-Kaplan Executive Function System Stuifbergen 2012

Subtotal (95% CI)

34

34

10.2 (2.1)

27

10.9 (2.7)


Total (95% CI)

56


-2

-1

Favours control


0

1

2


127

Analysis 4.8. Comparison 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up), Outcome 8 Visual functions. Review:


Comparison: 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up) Outcome: 8 Visual functions

Study or subgroup

Intervention


Control

N

Mean(SD)

N

Mean(SD)

10

10.6 (2.9)

9

10.4 (2.1)

Weight

IV,Fixed,95% CI


1 Intelligence Test-revised (HAWIE-R) Tesar 2005

Subtotal (95% CI)

10

23.9 %

0.07 [ -0.83, 0.98 ]

23.9 %

0.07 [ -0.83, 0.98 ]

76.1 %

0.10 [ -0.41, 0.60 ]

27

76.1 %

0.10 [ -0.41, 0.60 ]

36

100.0 %

0.09 [ -0.35, 0.53 ]

9

Heterogeneity: not applicable Test for overall effect: Z = 0.16 (P = 0.87) 2 Judgement of Line Orientation Test Stuifbergen 2012

Subtotal (95% CI)

34

27.8 (3.9)

34

27

27.4 (4.2)


Total (95% CI)

44


-0.5

-0.25

Favours control


0

0.25

0.5


128

Analysis 4.9. Comparison 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up), Outcome 9 Verbal functions. Review:


Comparison: 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up) Outcome: 9 Verbal functions

Study or subgroup

Intervention N

Mean Difference

Control Mean(SD)

N

Mean(SD)

Weight

IV,Fixed,95% CI


1 Controlled Oral Word Association Test (semantic fluency) Mäntynen 2013

58

25.5 (7.1)

40

24.2 (7.9)

78.2 %

1.30 [ -1.75, 4.35 ]

Stuifbergen 2012

34

36.1 (10.7)

27

36.4 (12)

21.8 %

-0.30 [ -6.08, 5.48 ]

Total (95% CI)

92

100.0 %

0.95 [ -1.75, 3.65 ]

67


-20

-10

Favours control


0

10

20


129

Analysis 4.10. Comparison 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up), Outcome 10 Everyday cognitive performance/patient’s report. Review:


Comparison: 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up) Outcome: 10 Everyday cognitive performance/patient’s report

Study or subgroup

Intervention


Control

N

Mean(SD)

N

Mean(SD)

66

10.1 (6.7)

143

10.7 (6.4)

Weight

IV,Random,95% CI


1 Memory Aids Questionnaire Lincoln 2002

Subtotal (95% CI)

66

143

15.9 %

-0.09 [ -0.38, 0.20 ]

15.9 %

-0.09 [ -0.38, 0.20 ]

6.4 %

-0.43 [ -1.18, 0.32 ]

6.4 %

-0.43 [ -1.18, 0.32 ]

15.9 %

0.01 [ -0.28, 0.31 ]

15.9 %

0.01 [ -0.28, 0.31 ]

16.1 %

-0.01 [ -0.30, 0.27 ]

16.1 %

-0.01 [ -0.30, 0.27 ]

10.3 %

0.17 [ -0.34, 0.68 ]

10.3 %

0.17 [ -0.34, 0.68 ]

Heterogeneity: not applicable Test for overall effect: Z = 0.62 (P = 0.54) 2 Memory Functioning Questionnaire/overall memory Chiaravalloti 2005

14

Subtotal (95% CI)

14

3.31 (1.25)

14

3.86 (1.23)

14

Heterogeneity: not applicable Test for overall effect: Z = 1.12 (P = 0.26) 3 Everyday Memory Questionnaire/patient’s report Lincoln 2002

Subtotal (95% CI)

66

22.4 (23.6)

66

142

22.1 (21.1)

142

Heterogeneity: not applicable Test for overall effect: Z = 0.09 (P = 0.93) 4 Dysexecutive Syndrome Questionnaire/patient’s report Lincoln 2002

Subtotal (95% CI)

70

20.4 (14.3)

70

145

20.6 (15.2)

145

Heterogeneity: not applicable Test for overall effect: Z = 0.09 (P = 0.93) 5 Strategy Subscale of the Multifactorial Memory Questionnaire Stuifbergen 2012

Subtotal (95% CI)

34

43.12 (11.93)

34

27

41.15 (10.65)

27

Heterogeneity: not applicable Test for overall effect: Z = 0.66 (P = 0.51) 6 Multiple Sclerosis Neuropsychological Questionnaire Mäntynen 2013

58

-24 (8.1)

40

-28.5 (10.5)

12.6 %

0.49 [ 0.08, 0.90 ]

Stuifbergen 2012

34

-28.41 (11.13)

27

-26.15 (11.56)

10.3 %

-0.20 [ -0.70, 0.31 ]

23.0 %

0.16 [ -0.51, 0.83 ]

Subtotal (95% CI)

92

67

Heterogeneity: Tau2 = 0.18; Chi2 = 4.25, df = 1 (P = 0.04); I2 =76%

-2

-1

Favours control

0

1

2


(Continued . . . )


130


Intervention N


Control Mean(SD)

N

Mean(SD)

-27.9 (11.7)

40

-36.8 (12.6)

Weight

IV,Random,95% CI


IV,Random,95% CI

Test for overall effect: Z = 0.47 (P = 0.63) 7 Perceived Deficits Questionnaire Mäntynen 2013

Subtotal (95% CI)

58

58

12.5 %

0.73 [ 0.32, 1.15 ]

40

12.5 %

0.73 [ 0.32, 1.15 ]

578

100.0 %

0.11 [ -0.12, 0.33 ]


Total (95% CI)

400


-2

-1

Favours control


0

1

2


131

Analysis 4.11. Comparison 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up), Outcome 11 Everyday cognitive performance/carer’s report. Review:


Comparison: 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up) Outcome: 11 Everyday cognitive performance/carer’s report

Study or subgroup

Intervention N


Control Mean(SD)

N

Mean(SD)

108

17.6 (18.7)

Weight

IV,Fixed,95% CI


1 Everyday Memory Questionnaire/carer’s report Lincoln 2002

Subtotal (95% CI)

55

18.3 (19.5)

55

37.5 %

0.04 [ -0.29, 0.36 ]

37.5 %

0.04 [ -0.29, 0.36 ]

38.2 %

-0.03 [ -0.36, 0.29 ]

38.2 %

-0.03 [ -0.36, 0.29 ]

24.3 %

0.13 [ -0.27, 0.53 ]

40

24.3 %

0.13 [ -0.27, 0.53 ]

258

100.0 %

0.03 [ -0.17, 0.23 ]

108

Heterogeneity: not applicable Test for overall effect: Z = 0.22 (P = 0.82) 2 Dysexecutive Syndrome Questionnaire/carer’s report Lincoln 2002

Subtotal (95% CI)

56

20.6 (16.5)

56

110

21.2 (17.4)

110

Heterogeneity: not applicable Test for overall effect: Z = 0.21 (P = 0.83) 3 Multiple Sclerosis Neuropsychological Questionnaire - Informant Mäntynen 2013

Subtotal (95% CI)

58

-19.3 (10.4)

58

40

-20.7 (11.1)


Total (95% CI)

169


-1

-0.5

Favours control


0

0.5

1


132

Analysis 4.12. Comparison 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up), Outcome 12 Depression. Review:


Comparison: 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up) Outcome: 12 Depression

Study or subgroup

Intervention

Mean Difference

Control

Weight

IV,Fixed,95% CI

Mean Difference

N

Mean(SD)

N

Mean(SD)

IV,Fixed,95% CI

Chiaravalloti 2005

14

-6.79 (8.15)

14

-7.29 (6.8)

14.5 %

0.50 [ -5.06, 6.06 ]

Mäntynen 2013

58

-9.8 (7.5)

40

-10 (6.2)

60.4 %

0.20 [ -2.52, 2.92 ]

Tesar 2005

10

-8.3 (5.8)

9

-8.3 (3.4)

25.1 %

0.0 [ -4.23, 4.23 ]

100.0 %

0.19 [ -1.92, 2.31 ]

1 Beck Depression Inventory

Total (95% CI)

82

63


-10

-5

Favours control


0

5

10


133

Analysis 4.13. Comparison 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up), Outcome 13 Quality of life. Review:


Comparison: 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up) Outcome: 13 Quality of life

Study or subgroup

Intervention N

Mean Difference

Control Mean(SD)

N

Mean(SD)

Weight

IV,Fixed,95% CI


1 Brief version of the World Health Organization Quality of Life/psychological total score Mäntynen 2013

58

Total (95% CI)

58

14 (2.9)

40

13.7 (2.2)

40

100.0 %

0.30 [ -0.71, 1.31 ]

100.0 %

0.30 [ -0.71, 1.31 ]


-4

-2

Favours control


0

2

4


134

Analysis 4.14. Comparison 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up), Outcome 14 Fatigue. Review:


Comparison: 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up) Outcome: 14 Fatigue

Study or subgroup

Intervention


Control

N

Mean(SD)

N

Mean(SD)

10

-41.8 (15.5)

9

-31.7 (18.8)

Weight

IV,Random,95% CI


1 Fatigue Impact Scale Tesar 2005

Subtotal (95% CI)

10

37.0 %

-0.56 [ -1.49, 0.36 ]

37.0 %

-0.56 [ -1.49, 0.36 ]

63.0 %

0.27 [ -0.14, 0.67 ]

40

63.0 %

0.27 [ -0.14, 0.67 ]

49

100.0 %

-0.04 [ -0.83, 0.75 ]

9

Heterogeneity: not applicable Test for overall effect: Z = 1.20 (P = 0.23) 2 Fatigue Scale for Motor and Cognitive Fatigue/total score Mäntynen 2013

Subtotal (95% CI)

58

58

-62.2 (19.3)

40

-67.1 (16.5)


Total (95% CI)

68


-2

-1

Favours control


0

1

2


135

Analysis 4.15. Comparison 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up), Outcome 15 Anxiety. Review:


Comparison: 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up) Outcome: 15 Anxiety

Study or subgroup

Intervention

Mean Difference

Control

N

Mean(SD)

N

Mean(SD)

-26.71 (9.61)

14

-32.79 (11)

Weight

IV,Fixed,95% CI


1 State Trait Anxiety Inventory/state Chiaravalloti 2005

14

Total (95% CI)

14

14

100.0 %

6.08 [ -1.57, 13.73 ]

100.0 %

6.08 [ -1.57, 13.73 ]


-10

-5

Favours control


0

5

10


136

Analysis 4.16. Comparison 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up), Outcome 16 Impact of the disease. Review:


Comparison: 4 Cognitive training combined with other neuropsychological rehabilitation methods versus any control (longitudinal follow-up) Outcome: 16 Impact of the disease

Study or subgroup

Intervention


Control

N

Mean(SD)

N

Mean(SD)

40

-27.1 (17.3)


Weight

IV,Fixed,95% CI

IV,Fixed,95% CI

1 Multiple Sclerosis Impact Scale/psychological total score Mäntynen 2013

Subtotal (95% CI)

58

-24.9 (18.2)

58

61.1 %

40

0.12 [ -0.28, 0.53 ]

61.1 % 0.12 [ -0.28, 0.53 ]

Heterogeneity: not applicable Test for overall effect: Z = 0.59 (P = 0.55) 2 Control subscale of the Multiple Sclerosis Self Efficacy Scale Stuifbergen 2012

Subtotal (95% CI)

34

557.72 (157.84)

34

27

38.9 %

534.26 (201.06)

0.13 [ -0.38, 0.64 ]

27

38.9 % 0.13 [ -0.38, 0.64 ]

67

100.0 % 0.13 [ -0.19, 0.44 ]


Total (95% CI)

92


-1

-0.5

0

Favours control

0.5

1


ADDITIONAL TABLES Table 1. Description of participants

Number of partici- Sex pants

Age in years Mean

Education in years Mean

Severity of disease, Duration of disEDSS ease, years Mean Mean

986 (966 MS partic- 70% women ipants 30% men 20 healthy controls)

44.6

12.3

3.2

14.0

EDSS: Expanded Disability Status Scale MS: multiple sclerosis


137

Table 2. Description of interventions

Cognitive training Number of studies (%)

Learning compensatory strategies Number of studies (%)

18 (90%) (in ad- 11 (55%) dition to training, learning compensatory strategies in 9 studies) Trained cognitive domains: memory, attention, visuospatial functions, executive functions, problem-solving skills)

Multimodal Number of ses- Duration of ses- Frequency neuropsychosions sions Mean logical rehabili- Mean Mean tation Number of studies (%) 2 (10%)

17

1 hour

Duration of intervention Mean

2 times per week 9.5 weeks

Table 3. Four most commonly used outcome measures in the 20 studies

Measure

Number of studies (%)*

Paced Auditory Serial Addition Test (PASAT)

10 (50%)

Symbol Digit Modalities Test (SDMT)

10 (50%)

Controlled Oral Word Association Test

9 (45%)

Beck Depression Inventory (BDI)

8 (42%)

*Other outcome measures were also used in all studies.

Table 4. Results of the study Jonsson 1993

Study

Main result

Other results

Jonsson 1993

Immediate follow-up Visual perception: the intervention group improved significantly compared to the control group (P < 0.004) Beck Depression Inventory (BDI): the intervention group improved significantly compared to the control group (P < 0.04)

Immediate follow-up In most cognitive tests, improvement due to learning and unspecific treatment effects Visuospatial memory: the intervention group improved compared to the control group (trend to significance P = 0.08)


138

Table 4. Results of the study Jonsson 1993

(Continued)

Longitudinal follow-up (6 months) Visuospatial memory: the intervention group improved significantly compared to the control group (P = 0.05) BDI: the intervention group improved significantly compared to the control group (P = 0.03)

Visuomotor speed: the control group improved compared to the intervention group (trend to significance P = 0.07) State-Trait Anxiety Inventory (STAI): no change Longitudinal follow-up (6 months) Visual perception: the intervention group improved compared to the control group (trend to significance P = 0. 09) Sum of the 11 cognitive tests: the intervention group improved compared to the control group (trend to significance P = 0.09)

Table 5. Results of the studies of Benedict 2000 and Mendoza 2001

Study

Main result

Other results

Benedict 2000

Modified Social Aggression Scale: the intervention group Beck Depression Inventory (BDI): both groups showed improved significantly compared to the control group (P modest improvement (not statistically significant); < 0.001) change scores did not differ between groups Hogan Empathy Scale (HES), NEO-Personality Inventory (NEO-PI): no change

Mendoza 2001

Beck Depression Inventory (BDI): significant improve- Activity level: participants in the intervention group parment in the intervention group compared to the control ticipated more often in voluntary unit activities than did group (P < 0.001) participants in the control group (trend to significance P < 0.06) Cognitive functions: no change

Table 6. Recommendations for future studies of neuropsychological rehabilitation

No

Recommendation

1

The use of comprehensive risk of bias (e.g. Higgins 2011) or quality assessment criteria (e.g. Van Tulder 2003) as guiding principles in preparing the research designs

2

Detailed reporting of the methods applied in the study (e.g. flow charts including exact number of patients)

3

Sufficient sample sizes

4

Objective baseline assessment of the cognitive status of the patients

5

Evaluation of treatment effects in cognitively homogeneous groups

6

Detailed reporting of the most essential disease variables


139

Table 6. Recommendations for future studies of neuropsychological rehabilitation

(Continued)

7

Determination of the aim of the intervention beforehand and measuring it with the primary measure

8

Detailed reporting of the content of the interventions

9

Detailed reporting of the basic statistics and outcome assessment timing

10

The use of outcome measures which more extensively reflect everyday functioning and the generalised effects of the interventions, thus enabling assessment of whether individual rehabilitation aims related to everyday functions have been achieved

11

Longitudinal follow-ups to evaluate the permanence of the treatment effects

APPENDICES Appendix 1. Keywords {cogniti\*} OR {awareness} OR {cognitive dissonance} OR {comprehension} OR {consciousness} OR {imagination} OR {intuition} OR {learn\*} OR {visual perception} OR {thinking} OR {psycholinguistics} OR {neuropsychol\*} OR {memor\*} OR {attenti\*} {language} OR {speech} OR {problem solving} OR {mathematics} OR {information process\*} OR {visuospat\*} OR {visuoconstruct\*} OR {execut\*} OR {metacognit\*} AND {rehabilit\*} OR {restitut\*} OR {remediat\*} OR {restorat\*} OR {retrain\*} OR {train\*} OR {recover\*} OR {treat\*} OR {guid\*} OR {instruct\*} OR {teach\*} OR {stimulat\*} OR {exerci\*} OR {strateg\*} OR {counsel\*} OR {therap\*} OR {intervent\*} OR {manage\*}

Appendix 2. PubMed

(1966 to 28 May 2013) 1

exp demyelinating autoimmune diseases, cns/ or exp multiple sclerosis/

2

(multiple sclerosis or ms).mp. [mp=title, original title, abstract, name of substance word, subject heading word, unique identifier]

3

1 or 2

4

exp cognition/ or exp awareness/ or exp cognitive dissonance/ or exp comprehension/ or exp consciousness/ or exp imagination/ or exp intuition/ or exp learning/ or exp visual perception/ or exp thinking/ or exp psycholinguistics/

5

Neuropsychology/


140

(Continued)

6

exp Memory/

7

Attention/

8

Language/

9

exp Speech/

10

Problem Solving/

11

exp Mathematics/

12

Awareness/

13

6 or 11 or 7 or 9 or 12 or 8 or 4 or 10 or 5

14

(cogniti* or neuropsychol* or memor* or learn* or attenti* or information process* or language or visuospat* or visuoconstruct* or problem solving or reason* or execut* or metacognit*).mp. [mp=title, original title, abstract, name of substance word, subject heading word, unique identifier]

15

13 or 14

16

exp Rehabilitation/

17

Remedial Teaching/

18

“Recovery of Function”/

19

Exercise/

20

exp Counseling/

21

18 or 19 or 16 or 17 or 20

22

(rehabilit* or restitut* or remediat* or restorat* or retrain* or train* or recover* or treat* or guid* or instruct* or teach* or stimulat* or exerci* or strateg* or counsel* or therap* or intervent* or manage*).mp. [mp= title, original title, abstract, name of substance word, subject heading word, unique identifier]

23

22 or 21

24

3 and 23 and 15

25

exp clinical trial/ or clinical trial, phase i/ or clinical trial, phase ii/ or clinical trial, phase iii/ or clinical trial, phase iv/ or controlled clinical trial/ or multicenter study/ or randomized controlled trial/

26

(randomi* control* trial* or randomi* clinical trial* or random allocation* or controlled clinical trial* or clinical trial* or experimental clinical trial* or quasi experimental clinical trial* or controlled trial* or randomi*ed trial* or clinical randomi*ed trial* or controlled stud* or crossover stud* or rct or crt).mp.


141

(Continued)

[mp=title, original title, abstract, name of substance word, subject heading word, unique identifier] 27

25 or 26

28

27 and 24

29

limit 28 to humans

exp = exploded; / = vocabulary term; mp = multiple (free-text search).

Appendix 3. EMBASE (1974 to 28 May 2013) ’crossover procedure’/exp OR ’double blind procedure’/exp OR ’single blind procedure’/exp OR ’controlled clinical trial’/de OR ’randomized controlled trial’/exp OR random*:ab,ti OR factorial*:ab,ti OR crossover:ab,ti OR ’clinical trial’/exp OR (cross:ab,ti AND over:ab,ti) OR placebo:ab,ti OR ’double blind’:ab,ti OR ’single blind’:ab,ti OR assign*:ab,ti OR allocat*:ab,ti OR volunteer*:ab,ti OR quasi experimental:ab,ti AND (’multiple sclerosis’/exp OR ’demyelinating disease’/de OR ’optic neuritis’/exp OR ’acute disseminated encephalomyelitis’/exp OR ’myelooptic neuropathy’/exp OR ’myelitis’/de OR ’multiple sclerosis’:ab,ti OR ’chronic progressive multiple sclerosis’:ab,ti OR ’progressive relapsing multiple sclerosis’:ab,ti OR ’secondary progressive multiple sclerosis’:ab,ti OR ’primary progressive multiple sclerosis’:ab,ti OR ’relapsing remitting multiple sclerosis’:ab,ti OR ’remitting-relapsing multiple sclerosis’:ab,ti OR ’acute relapsing multiple sclerosis’:ab,ti OR ’optic neurities’:ab,ti OR ’neuromyelitis optica’:ab,ti OR encephalomyelitis:ab,ti OR ’clinically isolated syndrome’:ab,ti OR ’transverse myelitis’:ab,ti OR ’devic disease’:ab,ti OR ’demyelinating disease’:ab,ti OR ’demyelinating disorder’:ab,ti OR ’acute disseminated encephalomyelitis’:ab,ti OR adem:ab,ti) NOT [medline]/lim AND (cogniti*:ab,ti OR neuropsychol*:ab,ti OR memor*:ab,ti OR learn*:ab,ti OR attenti*:ab,ti OR ’information process’:ab,ti OR language:ab,ti OR visuospat*:ab,ti OR visuoconstruct*:ab,ti OR ’problem solving’:ab,ti OR reason*:ab,ti OR execut*:ab,ti OR metacognit*:ab,ti) AND (’rehabilitation’/ de OR ’counseling’/exp OR (rehabilit*:ab,ti OR restitut*:ab,ti OR remediat*:ab,ti OR restorat*:ab,ti OR retrain*:ab,ti OR train*:ab,ti AND recover*:ab,ti OR treat*:ab,ti AND guid*:ab,ti AND instruct*:ab,ti) OR teach:ab,ti OR stimulat*:ab,ti OR exerci*:ab,ti OR strateg*:ab,ti OR counsel*:ab,ti OR therap*:ab,ti OR intervent*:ab,ti OR manage*:ab,ti) AND [humans]/lim AND [embase]/lim

Appendix 4. CINAHL

(1981 to 28 May 2013) 1

(MH “Multiple Sclerosis”) or (MH “Autoimmune Diseases+”)

2

TX Multiple sclerosis or ms

3

S1 or S2

4

(MH “Cognition+”) or (MH “Mental Processes+”)

5

(MH “Neuropsychology”)

6

(MH “Attention”)

7

(MH “Language+”)


142

(Continued)

8

(MH “Mathematics+”)

9

S4 or S5 or S6 or S7 or S8

10

TX cogniti* or neuropsychol* or memor* or learn* or attenti* or information process* or language or visuospat* or visuoconstruct* or problem solving or reason* or execut* or metacognit*

11

S9 or S10

12

(MH “Rehabilitation+”)

13

(MH “Recovery”)

14

(MH “Counseling+”)

15

S12 or S13 or S14

16

TX rehabilit* or restitut* or remediat* or restorat* or retrain* or train* or recover* or treat* or guid* or instruct* or teach* or stimulat* or exerci* or strateg* or counsel* or therap* or intervent* or manage*

17

S15 or S16

18

S3 and S11 and S17

19

(MH “Clinical Trials+”)

20

TX randomi* control* trial* or randomi* clinical trial* or random allocation* or controlled clinical trial* or clinical trial* or experimental clinical trial* or quasi experimentasl clinical trial or controlled trial* or randomi*ed trial* or clinical randomi*ed trial* or controlled stud* or crossover stud* or rct or crt

21

S19 or S20

22

S18 and S21

23

limiters: Human

24

S22 and S23


143

WHAT’S NEW Last assessed as up-to-date: 28 May 2013.

Date

Event

Description

28 May 2013

New search has been performed

New search. Six new trials have been included.

28 May 2013

New citation required and conclusions have changed

Conclusions changed.

CONTRIBUTIONS OF AUTHORS Both authors (ER and PH) participated in the whole review process: selection and evaluation of studies, extraction of data, analyses and writing the review.

DECLARATIONS OF INTEREST Both authors (ER and PH) acted as authors of one of the included studies (Mäntynen 2013).

SOURCES OF SUPPORT Internal sources • No sources of support supplied

External sources • Finnish National Insurance Institution, Finland. The study had financial support from the Finnish National Insurance Institution.

INDEX TERMS Medical Subject Headings (MeSH) Cognition Disorders [∗ rehabilitation]; Depression [rehabilitation]; Memory Disorders [∗ rehabilitation]; Multiple Sclerosis [∗ psychology; rehabilitation]; Neuropsychology; Quality of Life; Randomized Controlled Trials as Topic


144

MeSH check words Humans


145

Neuropsychological aspects of multiple sclerosis.

Neuropsychological stability in multiple sclerosis.

Multiple sclerosis rehabilitation.

Multiple sclerosis rehabilitation.

The use of goal attainment scaling in neuropsychological rehabilitation in multiple sclerosis.

Predictors and impact of the working alliance in the neuropsychological rehabilitation of patients with multiple sclerosis.

Future directions of multiple sclerosis rehabilitation research.

Update on rehabilitation in multiple sclerosis.

Let's rehabilitate cognitive rehabilitation in multiple sclerosis.

Neuropsychological performance, brain imaging, and driving violations in multiple sclerosis.

Neuropsychological correlates of multiple sclerosis across the lifespan.

Cognition, cognitive dysfunction, and cognitive rehabilitation in multiple sclerosis.

Rehabilitation of a patient severely involved with multiple sclerosis.

Cognitive rehabilitation in multiple sclerosis: A systematic review.

Cognitive rehabilitation in multiple sclerosis: a randomized controlled trial.

Medical, functional, and legal aspects of vocational rehabilitation for people with multiple sclerosis.

A new cognitive rehabilitation programme for patients with multiple sclerosis: the 'MS-line! Project'.

Adaptive personalized training games for individual and collaborative rehabilitation of people with multiple sclerosis.

Alemtuzumab for Multiple Sclerosis.

Can we define a rehabilitation strategy for cognitive impairment in progressive multiple sclerosis? A critical appraisal.

Psychotherapy for multiple sclerosis.

Evaluating Neurocognitive Deficits in Patients With Multiple Sclerosis Via a Brief Neuropsychological Approach.

Neuropsychological features in childhood and juvenile multiple sclerosis: five-year follow-up.

Driving competences and neuropsychological factors associated to driving counseling in multiple sclerosis.