Study reference

Study characteristics

Patient characteristics ²

Intervention (I)

Comparison / control (C) ³

Follow-up

Outcome measures and effect size ⁴

Comments

Aparicio-Lopez, 2016

Type of study:

RCT

Setting and country:

Spain

Funding:

Not reported

Conflicts of interest:

None declared.

Inclusion criteria:

(1) suffering a right-hemisphere stroke diagnosed by Computed Axial Tomography or Magnetic Resonance Imaging; (2) being 18 or older at the time of lesion; (3) right-handedness; (4) Spanish as mother tongue; and (5) obtaining scores suggestive of VSN in the neuropsychological exploration protocol used to assess visuo-spatial attention.

Exclusion criteria:

(1) severe language alteration limiting the patient’s communicative

ability; (2) significant visual acuity impairment

caused by cataracts, diabetes, retinopathy and/or glaucoma; and (3) pre-morbid history of other neurological diseases, psychiatric disorders and/or drug abuse.

N total at baseline:

Intervention: 13

Control: 15

Important prognostic factors²:

For example

age ± SD:

I: 50.5 ±7.6

C: 45.7 ±8.1

Sex:

I: 61.5% M

C: 60% M

Groups comparable at baseline?

yes

The combined treatment group carried out the same cognitive treatment as the single treatment group, combined with the right hemifield eye-patching (RHEP). RHEP was implemented by using non-prescription glasses especially made for the study. These glasses had a completely opaque right half-field for each eye. This group wore these glasses during all cognitive

treatment sessions. Sessions lasted for one hour with on average 3.4 sessions per week and 15 sessions in total.

The single treatment group followed a cognitive rehabilitation program

using the computer-based platform Guttman-

NeuroPersonalTrainer. Exercises included attention,

memory and executive function tasks. In all cases, adequate performance of the assigned tasks required visual processing of stimuli homogeneously distributed across the screen. Sessions lasted for one hour with on average 2.9 sessions per week and 15 sessions in total.

Length of follow-up:

Not reported

Loss-to-follow-up:

1 person was lost post-randomisation but unclear which group that person had been allocated to.

Incomplete outcome data:

Not reported

ADL: Catherine Bergego Scale

A measure of visuo-spatial neglect related daily functioning, was completed by either the patient or a relative. It is unclear how many participants completed each version. Presented are means and SD; within-group differences were tested using the singed rank test. Between-group differences in pre-to-post changes were tested using the Mann-Withney U test.

*patient version:

I:

pre=6.8 ±6.9

post=4.4 ±3.4

within-group difference: p=0.06

C:

pre=7.1 ±6.5

post=6.1 ±6.1

within-group difference: p=0.93

Between-group difference: p=0.25

*relative version:

I:

pre=11.9 ±7.7

post=10.3 ±6.6

within-group difference: p=0.55

C:

pre=10.6 ±8.5

post=8.8 ±6.7

within-group difference: p=0.31

Between-group difference: p=0.75

Neglect: line bisection

Mean (SD) percent of positively for rightward deviations:

I:

Pre: 27.51 ±16.7

Post: 21.13 ±16.39

Within-group difference: p=0.02

C:

Pre: 20.58 ±10.39

Post:11.28 ±6.13

Within-group difference: p=0.02

Between0group difference: p=0.39

Neglect: target cancellation

Bell cancellation test, mean (SD) test score:

I:

Pre: 14.08 ±9.35

Post: 23.15 ±9.20

Within-group difference: p=0.003

C:

Pre: 20.07 ±9.16

Post: 29.67 ±3.83

Within-group difference: p=0.001

Between-group difference:p=0.86

Vergelijking: interventie versus controle

P=stroke + negect

I= eye patching + cognitive training

C=cognitive training (controle)

O=ADL, neglect (line bisection, target cancellation)

Kerkhoff, 2014

Type of study:

RCT

Setting and country:

Germany

Funding:

No financial support received.

Conflicts of interest:

None declared.

Inclusion criteria:

a single right-hemisphere stroke,

visual neglect in 2 screening tests (horizontal line bisection of a 200 mm long × 10 mm broad black line, number cancellation), and the ability to execute daily neglect training for

30 minutes. Neglect was diagnosed when bisection deviated >5 mm to the right side and when >1 omissions were found on the left side than on the right side of the cancellation test.

Exclusion criteria:

psychiatric, ophthalmological, or other neurological diseases.

N total at baseline:

Intervention: 12

Control: 12

Important prognostic factors²:

Age (yrs) ± SD:

I: 64 ±3

C: 64 ±3

Sex:

I: 58.3% M

C: 66.7% M

Groups comparable at baseline?

Yes

Smooth pursuit eye movement training (I1)

During smooth pursuit eye-movement training (SPT),

Participants watched random displays of 20 to 60

identically colored and sized squares, moving horizontally from right to left at constant velocity, providing a coherent

stimulus pattern. The

color, size (1°-2.5°), and velocity (3.1-12.6 degrees/s) of the stimuli were varied from time to time. The patient was encouraged to conduct smooth pursuit eye movements, repeatedly following the stimulus pattern from right to left,

without head movements. Participants were sitting

upright in the bed

Stimuli were displayed on a laptop with a 15-inch monitor, centrally

placed at a viewing distance of 0.40 m (39.2° versus 30.1°

horizontal/vertical size) in front of the patient on a board seated solidly across the patient’s bed.

The therapist watched continuously from the side whether the patient made smooth pursuit eye movements to the contralesional side. If patients did not execute the appropriate eye movements, the instructions were repeated. If patients were tired, a short break (2 minutes) was given, adjusted to patients’ individual requirements.

Patients received 20 treatment sessions (30 minutes each) in total (1 session daily, from Monday to Friday) over a period of 1 month

Visual scanning training (I2)

During visual scanning training (VST), patients viewed stationary displays of stimuli implemented in the same software as SPT. Stimuli were geometrical objects (circles, triangles,

squares, etc), stars, letters, or digits. The number of stimuli (5-60), their size

(1°-2.5°), color, and spatial arrangement were varied (systematic versus. unordered, with/without line numbering, with/without subsidiary horizontal reference lines, and with/without a red vertical “anchor line” on the leftmost side of the display). Patients were trained to scan systematically from left to right and from top to bottom, naming all objects, or counting certain stimuli.

Participants were sitting

upright in the bed

Stimuli were displayed on a laptop with a 15-inch monitor, centrally

placed at a viewing distance of 0.40 m (39.2° versus 30.1°

horizontal/vertical size) in front of the patient on a board seated solidly across the patient’s bed.

The therapist watched continuously from the side whether the patient made saccadic eye movements. If patients did not execute the appropriate eye movements, the instructions were repeated. If patients were tired, a short break (2 minutes) was given, adjusted to patients’ individual requirements.

Patients received 20 treatment sessions (30 minutes each) in total (1 session daily, from Monday to Friday) over a period of 1 month

Length of follow-up:

1 month

Loss-to-follow-up:

Intervention:

N (%): 0

Control:

N (%): 0

Incomplete outcome data:

Not reported

Data were presented as means or mean differences and within or between-group differences were tested using ANOVA.

Neglect: line bisection

Horizontal stick bisection test : a total of 3 visual and 3 tactile trials were performed of indicating the midpoint of a wooden bar. The mean across trials was transformed into a score (range 0-3) reflecting magnitude of deviation.

Baseline-posttest mean difference:

I1: 5.16, p<0.001

I2: 2.83, p=0.041

ADL: Help Index

Measured with the Help Index (performance on 8 ADLs rated by a nurse on a 5-point scale; range 8-40, higher scores indicate poorer functioning).

Pre- and post-intervention data were presented graphically only.

Between-group differences:

Time-by-group interaction p=0.45

ADL: Barthel Index

Measured with the Help Index (performance on 10 ADL and mobility tasks; range x-x, higher scores indicate better functioning).

Presented are means ±SD, group differences over time are tested using ANOVA.

I1:

Pre=15 ±5

Post=26 ±8

I2:

Pre=11 ±4

Post=28 ±5

Between-group differences: Time-by-group interaction p=0.16

Vergelijking: interventie versus andere interventie

P=stroke + neglect

I1=smooth pursuit eye movement

I2=visual scanning

O=neglect, ADL

Keller, 2010

Type of study:

RCT

Setting and country:

Germany

Funding:

None received

Conflicts of interest:

None declared

Inclusion criteria:

-unilateral brain lesions of the occipital lobe confirmed by MRI

-Because patients with quadrantanopia can be affected in everyday life as severely as hemianopic patients, both types of visual field

deficits were included in the study.

-normal hearing, as measured by audiometry in each ear, with no signs of asymmetry between ears, as well as a normal or corrected binocular visual acuity for near and far space.

Exclusion criteria:

-participants with visual neglect or signs of aphasia were excluded from the study.

N total at baseline:

Intervention: 10

Control: 10

Important prognostic factors²:

age ± SD:

I: range 16-85

C: range 43-83

Sex:

I: 60% M

C: 60% M

Groups comparable at baseline?

Yes

Audiovisual stimulation (I1):

For the presentation of the visual and acoustic stimuli, 48 red LED and piezoelectric

loudspeakers were positioned in 3 rows at 8, 16, 24, 32, 40, 48, 56, and 64 on either side of a central fixation Point. Sounds of 80 dB were

created by a white-noise generator. Visual and acoustic stimuli were presented synchronously in the same spatial positions

with a duration of 100 milliseconds. All training sessions were conducted in a dimly illuminated room. The patients sat in chairs 60 cm in front of the pparatus, with their eyes adjusted to the center of the apparatus. For patients with homonymous hemianopia, 70% of all stimuli were presented in the blind visual field. For patients with quadrantanopia, 70% of the

stimuli were presented in the 2 upper or 2 lower rows of the affected side. Stimuli to the blind visual field and intact visual field were presented in random sequence. Patients were

instructed to detect the presence of visual targets by pressing a response button as fast as possible. To prevent patients from

reacting to false positives, 20% catch trials with solely acoustic stimulation were implemented in each training session. Whenever

patients responded to a catch trial, a computer-generated nickering of a horse sounded. For each training session, patients were explicitly instructed to execute eye movements with their head held straight forward. The training was carried out under binocular conditions. Whenever a patient complained of tiredness, the training was interrupted for 1 minute.

Each patient received 20 therapy sessions (each session lasting 30 minutes) over 3 weeks.

Visual stimulation (I2):

For the visual exploration training, we used the same apparatus as described before. Settings of stimuli presentation were identical as for the AVT, with the exception that the sound was turned off during all

training sessions and catch trials were not needed.

Each patient received 20 therapy sessions (each session lasting 30 minutes) over 3 weeks.

Length of follow-up:

3 weeks

Loss-to-follow-up:

Not reported

Incomplete outcome data:

Not reported

ADL: Help Index

Modified version of the Help Index - 5 items were selected from a questionnaire that can be observed in an in-patient rehabilitation setting.

Within-group differences:

I1: p=0.03

C: p=0.02

Mean post-test scores:

I1: 1.5

C: 5.0

Between-group differences at post-test (ANOVA): p=0.04

Vergelijking: interventie versus een controle

P=stroke (18), tumour (1), trauma (1) with hemianopsia

I1=audio-visual stimulation

C=control (as the same visual stimulation was received as in the intervention group)

O=ADL

Modden, 2012

Type of study:

RCT

Setting and country:

Germany

Funding:

None

Conflicts of interest:

None declared

Inclusion criteria:

-homonymous hemianopia with a posterior cerebral

artery stroke

Exclusion criteria:

-visual neglect, eye-movement disorders, neuropsychological

disorders like aphasia, dysexecutive syndromes,

memory deficits, or higher order motor impairments like

apraxia

N total at baseline:

Intervention 1: 15

Intervention 2: 15

Control: 15

Important prognostic factors²:

age ± SD:

I1: 58.3 ±11.4

I2: 57.1 ±8.3

C: 59.0 ±11.1

Sex:

I1: 66.7% M

I2: 60% M

C: 46.7% M

Groups comparable at baseline?

yes

Restitution therapy (I1)

Compensatory therapy (I2)

I1: A therapy-integrated perimeter program

(provided by Teltra company) created the exact

measurement of the individual visual field border. Using that measurement, a series of coloured targets appeared on a blue screen anywhere at 1 of 10 positions on the border line. A randomly presented first fixation target (a rotating arrow) announced the second stimulus target in the hemianopic border zone (basic principle of covert attention shift). The patients were instructed to respond (by pressing a key) to each stimulus target (colored and flickering frames, beams, and spots) as soon it was perceived. The program contained no adaptive difficulty levels. Eye movements were not allowed, and this was controlled by the assistant.

I2: The “Exploration” task (from RehaCom, provided by HASOMED GmbH, Magdeburg, Germany) was adapted individually according to the side of the hemianopia. On a dark background, different bright

stimuli arranged in rows and columns were presented. A ring (diameter of 2 cm) moved line by line (interlaced) on a matrix unit over the field. The participant was instructed to follow the ring (starting point to an outmost fixation in the blind side) by eye movements and to identify a critical targeted icon. The targets were not always distributed homogeneously

but were clustered in the blind side. Thus, the

exploration in the hemianopic field was further promoted. The patients had to respond (by pressing a key) when the

targeted icon was perceived in the circle. The program contained several difficulty levels. In levels 1 to 20, all lines were completely filled with symbols, whereas there were omissions in the rows of symbols in levels 21 to 30 to increase the difficulty.

The daily duration of each session for all groups was 30 minutes, and each participant received a total of 15 sessions.

Occupational therapy (C)

After a standardized assessment of daily living activities, the therapy consisted of individually

adapted stimulation of daily activity tasks to compensate

via eye-, head-, and body movements. These compensation strategies included aspects of spatial and body perception,

searching or arranging objects, pen and paper searching task, reading maps or newspapers, and self-care activities. The participant was instructed to perform systematic eye movements toward the lost visual field. The interventions were carried out in the treatment rooms, on the wards, in a

kitchen or a bathroom, outside in the park, or in a supermarket. Patients receiving RT and CT did not receive OT in the context of their standard rehabilitation treatment.

The daily duration of each session for all groups was 30 minutes, and each participant received a total of 15 sessions.

Length of follow-up:

?

Loss-to-follow-up:

None were lost to follow-up, but 2 participants (1 in I1, 1 in I2) were unable to receive the allocated intervention due to health impairments. These participants were excluded from the analyses and replaced by the next recruited patient.

Incomplete outcome data:

Not reported

ADL functioning: Barthel Index (extended)

Within-group difference in mean scores ±SD:

I1: 1.5 ±2.8, p=0.03

I2: 3.3 ±3.6, p=0.005

C: 1.8 ±2.0, p=0.03

Between-group differences:

-I1 versus C: not significant

-I2 versus C: not significant

-I1 versus I2: not significant

Perimetry: Test Battery of Attentional Performance

Within-group difference in mean scores ±SD:

I1: 3.9 ±4.9, p=0.003

I2: 2.9 ±4.0, p=0.01

C: 1.3 ±4.7, p=0.32

Between-group differences:

-I1 versus C: not significant

-I2 versus C: not significant

-I1 versus I2: not significant

Neglect: cancellation task

Mean (SD) number of omissions:

I1 5.3 ±10.5, p=0.005

I2: 5.4 ±5.2, p=0.003

C: 2.3 ±5.0, p=n.s.

Between-group differences:

I1 versus C: n.s.

I2 versus C: p=0.048

I1 versus I2: n.s.

Vergelijking: intervention versus intervention and intervention versus control

P= stroke

I1: restitution à visual stimulation

I2: compensatory à visual scanning

C: standard occupational therapy (usual care)

O= ADL function, visual field (perimetry), neglect

Crotty, 2018

Type of study:

RCT

Setting and country:

Australia

Funding:

Not reported

Conflicts of interest:

AH participated in the development of the device.

Inclusion criteria:

-experienced a stroke between

two weeks and six months prior; had homonymous

hemianopia (any field loss reached above and below

the horizontal axis, as measured by Medmont C-100 perimetry performed by a

neuro-ophthalmologist);

-scored ≥25 on the MMSE;

-corrected vision of at least 6/18;

-able to mobilise a distance of 35 metres independently or

with supervision or standby assistance; and ≥18 years of age.

Exclusion criteria:

-unable to follow instructions due to aphasia or poor English language skills

N total at baseline:

Intervention: 11

Control: 13

Important prognostic factors²:

age ± SD:

I: 68.6 ±13.1

C: 60.7 ±17.2

Sex:

I: 46.2% M

C: 63.6% M

Groups comparable at baseline?

yes

Scanning training

Participants received the

standardized program which consisted of three weeks (9 sessions) static scanning training using the NVT scanning device and four weeks (12 sessions) of

mobility training using NVT scanning techniques. The NVT scanning device teaches a pattern of

systematic visual search strategies, defining the

perimeter of the affected visual field as a starting

point for consistent visual search patterns. A standardized presentation of a panel of colored lights

requires the patient to utilize both head and eye

movement towards the affected visual field, with patterns of light that develop and improve speed and accuracy of visual scanning, speed of visual processing, visual memory and visual spatial reasoning.

The program then transfers visual scanning strategies,

into the ‘real life’ activities for strategies to become automatic, graded from a structured familiar

environment to more dynamic busy environments,

including shopping malls and busy roads, requiring high level scanning skills, complex visual processing

and problem solving.

Usual care

Participants allocated to the control group received usual care, including one-to-one occupational therapy and mobility instruction promoting visual scanning and mobility training by the Guide Dogs Association SA/NT. Scanning training included a panel of coloured lights. Therapy was individually targeted and dosage/type of therapy was determined and recorded by the treating therapist. This group functioned as a waiting-list control group, with the standardized training using the NVT protocol offered at the completion of the three month follow-up.

Length of follow-up:

7 weeks

Loss-to-follow-up:

No loss to follow-up at 7 weeks. Though participants were lost during the post-intervention follow-up.

Intervention:

N (%): 1 (9.1%)

Reasons (describe)

-deceased (1)

Control:

N (%): 3 (23.1%)

Reasons (describe)

-deceased (1)

-refused full assessment (1)

-family circumstances (1)

Incomplete outcome data:

-4 cases with incomplete data due to missing the 3-month assessment. These participants were excluded from the analyses.

Neglect: Behavioural Inattention Test

Mean (±SD) pre-post differences in total scores:

I: -7.8 ±15.8

C: -5.3 ±7.5

Between-group mean difference: 2.45 (95%CI -8.9, 13.8), ANOVA p=0.94

Vergelijking: intervention versus control

P=stroke

I=visual scanning training

C=usual care

O=neglect,

Rowe, 2017

Type of study:

RCT

Setting and country:

United Kingdom

Funding:

Non-government.

Conflicts of interest:

None declared

Inclusion criteria:

-18 years of age or older;

-Best corrected visual acuity of 0.5 or better in each eye at distance;

-Stable homonymous hemianopia (partial or complete) induced by recent stroke, defined following WHO guidelines, present over 2 weeks (to exclude rapid recovery cases) but less than 26 weeks

prior to randomization;

-Refractive error within ±5 dioptres;

-Willing and able to give consent for the study;

-Prior to stroke able to read and understand English.

Exclusion criteria:

-unable to consent due to severe cognitive impairment;

-assessed to have ocular motility impairment and/or visual inattention

in addition to the visual field impairment;

-had pre-existent visual field impairment due to previous stroke.

N total at baseline:

I1: 27

I2: 30

C: 30

Important prognostic factors²:

age ± SD:

I1: 69.9 ±12.9

I2: 70.9 ±11.2

C: 66.2 ±11.3

Sex:

I1: 84.6% M

I2; 56.7% M

C: 69.0% M

Groups comparable at baseline?

Yes

Prisms (I1)

Visual scanning training (I2)

I1: Participants were given sector Fresnel prisms of 40

prism dioptre strength on their glasses (or plain glasses if not already

worn). Separate prism segments were used as a mechanical displacement

to expand the upper and lower quadrants. Full fitting details are detailed in the protocol. Participants were advised to wear the

prisms for a minimum of 2 hours daily, for a minimum 6 weeks, from prism affixation; after this, they could elect to continue treatment if wished.

I2: Participants were assessed and provided with visual search training.

This comprised an A4 landscape card with horizontal and diagonal

numbered circles radiating out from a central fixation target. Full instructions for training are detailed in the protocol.13 Participants

were instructed to continually scan between the various targets for 30 minutes daily for a minimum 6 weeks, after which they could elect

to continue treatment if they wished. Participants were instructed on the search exercises to ensure their understanding of doing this training. In addition, printed instructions were provided with the visual training target cards.

Standard care (C)

Participants were given information leaflets from the UK Stroke

Association and the UK Royal National Institute for the Blind about visual impairment following stroke. Participants in all treatment groups received these information leaflets.

Length of follow-up:

6 weeks

Loss-to-follow-up:

Intervention:

I1 N (%): 3 (11.1%)

I2: N (%): 5 (16.7%)

Reasons (describe)

-withdrawal (1/0)

-partial withdrawal (1/5)

-loss to follow-up (1/0)

Control:

N (%): 8 (26.7%)

Reasons (describe)

-withdrawal (1)

-partial withdrawal (3)

-loss to follow-up (4)

Incomplete outcome data:

Intervention:

I1: N (%): 2

I2: N (%): 5

Reasons (describe)

-perimetry not completed

Control:

N (%): 5

Reasons (describe)

-perimetry not completed

Visual field: Goldmann perimetry and Humphrey or Octobpus perimetry

Within-group mean difference (95%CI):

I1:0.052 (-0.006, 0.111)

I2: 0.082 (0.019, 0.1444)

C: 0.035 (-0.031, 0.1018)

Between-group difference (ANOVA): F=0.59, p=0.56

ADL functioning: Nothingham expanded activities of daily living assessment (NEADL)

Data not reported.

Within-group differences: ANOVA: p=0.93

Vergelijking: interventie versus andere interventie; interventie versus control

P=stroke

I1=prisms

I2=visual scanning training

C=usual care

O=visual field (perimetry), ADL functioning

Elshout, 2016

Type of study:

RCT-cross-over

Setting and country:

Netherlands

Funding:

Government; Industry

Conflicts of interest:

Not reported

Inclusion criteria:

-Patients in the chronic phase of stroke (>10 months post CVA);

-no signs of visual neglect (line bisection test)

-aged 18-75 years;

- able to undergo (f)MRI scanning

- macular sparing of at least 2° on the Goldmann perimetry test

Exclusion criteria:

-

N total at baseline:

Cross-over design: n=30

Important prognostic factors²:

age ± SD:

51.2 (range 29-74)

Sex:

81.5% M

Groups comparable at baseline?

Not applicable

Defect training

Each patient received a training unit at home to create a controlled training environment. This unit consists of a container, to be placed on a table, with a top cover and side covers to present a dark visual surround for the training stimuli with the exclusion of stray light. Mini Mac computer, keyboard, and mouse, a support with 24’’ LED monitor, webcam, and chin/head rest were positioned inside the matte black container. Viewing distance was fixed at 40 cm. The subject’s face was indirectly lighted with a TL light for eye tracking with the webcam. The computer was prepared with eye tracking software, and training programs that were adjusted to the particular visual field defect of the patient.

During defect training, high contrast stimuli(C>0.9) were offered within the field defect along its border.

Fifteen patients were trained with a static white point stimulus on a black screen. The other group of 15 patients received an optic flow-discontinuity stimulus.

The length of one training session was on average 12 min (depending on the number of trials set per session and amount of fixation errors). The number of trials in a session ranged between 60 and 100, depending on the shape and quality of the visual field defect. The stimuli were randomly presented for each session. The patients trained 1 h a day, 5 days a week during 8 weeks to complete at least 40 h of training per hemifield.

Intact training (placebo/sham)

Same set up as in the Intervention group, but stimuli were presented within the intact field at about the same eccentricities as for the patient’s defect training.

Length of follow-up:

8 weeks

Loss-to-follow-up:

N (%): 3 (10%)

Reasons (describe):

-dropped out during first training round for personal reasons (3)

Incomplete outcome data:

-Complex visual field defect precluded definition of the Goldmann isopter (2)

-At least one of the three Humphrey measurements was unreliable (5)

Field of view: perimetry

• Goldmann

Comparing defect training with intact training :

Within-group mean difference:

I: 4.34, SE=1.03

C: 0.82, SE=0.84

Between-group difference: p=0.045

• Humphrey

Comparing defect training with intact training :

Between-group mean difference: not reported

Vergelijking: intervention versus control

P=stroke

I=visual stimulation

C=placebo/sham

O=field of view (perimetry)

De Haan, 2015

Type of study:

RCT

Setting and country:

Netherlands

Funding:

Government, University, non-government

Conflicts of interest:

None declared

Inclusion criteria:

-presence of a HVFD, at least a quadrantanopia, restricted to one half of the visual field, due to acquired post-chiasmatic brain injury

-time since onset

had to exceed 5 months

Exclusion criteria:

-ocular diseases affecting the visual field or binocular visual acuity, signs

of severe physical impairments or (neuro)psychological disorders.

-Neglect based on the Balloons, drawings, line bisection and Rey complex figure test

N total at baseline:

Intervention: 26

Control: 23

Important prognostic factors²:

age ± SD:

I: 55 ±10.1

C: 57 ±13.0

Sex:

I: 69.3% M

C: 60.9% M

Groups comparable at baseline?

No: difference in visual field

Scanning training

Training was given by occupational therapists. The training consisted of 15 individual sessions of 60–90 minutes each, 18.5

hours of face-to-face training in total during a period of 10 weeks. The aim is to teach patients with HVFD to apply a systematic, anticipatory scanning strategy in order to

compensate for their visual field defect during a wide range of mobility-related activities. Patients are taught a scanning strategy consisting of a triad of horizontal saccades. In the IH-CST, the scanning strategy is to start with one large saccade towards the blind side, followed by a large saccade ending on the peri-central seeing side,

and then back to the starting point of looking straight forward. Patients learn to generate this scanning rhythm endogenously on an anticipatory basis and to adjust the speed of repetition of this scanning rhythm to environmental demands and to the speed of walking, cycling, etc.

The protocol starts with exercises for improving awareness of the size and shape of the visual field defect and its consequences for daily life activities. Then, the scanning rhythm is systematically practiced with several exercises

gradually increasing speed and amplitude of the scanning triad. At first, only eye movements are allowed, since eye movements are faster than head movements, they do not lead to neck muscle

complaints and they naturally precede head movements. In a later stage, head movements

following the eye movements are practiced to increase the range of scanning. A substantial

part of the training is dedicated to the practicing of the scanning rhythm in a range of daily life mobility situations, with increasing complexity and cognitive load, in order to optimize

transfer to visual activities and participation in daily life. For every exercise, specific targets are

defined for speed and amplitude of the scanning rhythm as well as transfer to an activity of daily life, which must be reached before the participant proceeds to the next exercise. For example, practicing the scanning rhythm during cycling will only be started after scanning during walking is performed without problems. Depending on the needs of the patient, some other compensatory techniques are practiced, for example searching for an object on a shelf. The

main reason for including these exercises is that they are expected to increase insight into the field defect. The focus of the IH-CST, however, is on applying a systematic, anticipatory scanning rhythm during a wide range of mobility-related activities.

In order to stimulate transfer to daily life, homework assignments

are included in the training protocol. The first homework assignment is aimed at improving insight in the visual field defect. In this assignment, the patient has to answer a number of questions on what they see and cannot see when looking straight forward in a number of predefined situations. Further homework assignments are aimed at practicing the scanning rhythm in daily life situations, stimulating transfer to daily life. Homework starts with practicing the scanning

rhythm using pieces of paper three times a day for five minutes. First a smaller band is used and then a wider band, in order to increase the amplitude of the saccades.

Waist list control

Length of follow-up:

10 weeks

Loss-to-follow-up:

Intervention:

N (%): 2 (6.7%)

Reasons (describe)

-low compliance (2)

-expanding brain tumour (1)

-time between T1 and T2 was too long (1)

Control:

N (%): 1 (4.2%)

Reasons (describe)

- time between T1 and T2 was too long (1)

Incomplete outcome data:

Complete data for outcomes of interest (for those who were not lost to follow-up)

Visual field: Goldmann perimetry (expressed as Functional Field Score)

I:

Pre: 57.9 ±7.80

Post: 57.75 ±6.74

Within-group difference: n.s.

C:

Pre: 63.79 ±11.41

Post: 62.58 ±11.13

Within-group difference: n.s.

Between-group difference: n.s.

GLM repeated measures: interaction time-by-group: n.s.

Visual functioning, Cerebral Visual Disorders questionnaire (CVD) includes

Measures items from the Unawareness and behavioural neglect index (UBNI), which measures impact of neglect on daily activities, and items asking about difficulty experienced in twelve specific activities. Higher scores indicate more difficulty.

Mean (± SD):

I:

Pre: 0.44 ±0.16

Post: 0.36 ±0.13

Within-group difference: p<0.05

C:

Pre: 0.45 ±0.15

Post: 0.46 ±0.16

Within-group difference: n.s.

Between-group difference: p<0.05

GLM repeated measures: interaction time-by-group: p=<0.05

Vergelijking: intervention versus control

P=stroke (41), TBI (3) PHT (1), AVM extirpation (1), combination (3)

I=visual scanning training

C=control

O=visual field (perimetry), visual functioning

Schuett, 2012

Type of study:

RCT cross-over

Setting and country:

Germany

Funding:

Government

Conflicts of interest:

Not reported

Inclusion criteria:

-left- or right-sided homonymous

visual field loss

-no prior treatment for their visual field defect

-native German speakers

-at least 5 years of education

Exclusion criteria:

-reduced visual acuity (50.90 for near and far binocular vision), impaired spatial contrast sensitivity (Vistech contrast sensitivity test, 1988), visual adaptation, disturbances of the anterior visual pathways or of the oculomotor system, macular disease (according to ophthalmologic examination), aphasia, premorbid reading disorders,

pure alexia (vertical word reading test; Zihl, 1995a, 2011),

impairments of visual–lexical numerical processing (horizontal and vertical number reading; Zihl, 1995a, 2011), or verbal memory deficits

-None of the patients had visual neglect as assessed

by tests in accordance with the Behavioural Inattention Test

N total at baseline:

Intervention: 18

Control: 18

Important prognostic factors²:

age ± SD:

I: 64.0 ±11.1

C: 63.7 ±13.3

Sex:

I: 83.3% M

C: 83.3% M

Groups comparable at baseline?

Yes

Visual scanning training (I1)

For improving visual exploration, we used standardized versions of the

visual search paradigm (parallel and serial search mode), which have proven to be useful in this regard (Zihl, 1995b, 2011). Patients were systematically trained to use larger saccadic eye movements. Training material consisted of visual search displays extending 50° horizontally and 42° vertically. We used different target and distractor letters of varying similarity as stimuli. Each training trial was composed of the presentation of a visual search display. Patients were instructed to fixate on a cross in the centre of the monitor and to search, after its offset, for a single target letter (e.g. ‘T’) among distractor letters (e.g. ‘O’s) as accurately and quickly as possible. In target-present trials, the patient was asked to press the left mouse

button, in target-absent trials, the right mouse button. Presentation

and, thus, visual search time was unlimited. This paradigm allows visual exploration or scanning eye movements to be trained and reinforced by the patient’s normal internal visual feedback and feedback given by the experimenter on visual search strategy, time and errors. Visual exploration

training started with the easiest, i.e. parallel search condition and progressed,

via a so-called mixed search condition to visual exploration training using more difficult serial visual search tasks. In addition to varying letter similarity during the course of training, visual search difficulty was also systematically increased by

increasing the number of stimuli. This training protocol was adjusted

to individual reading performance and training progress.

An individual training session lasted ~45 min and consisted of 10 practice units (30 trials each).

The treatment was administered and supervised by the experimenter, who sat beside the patient to give

verbal feedback on reading or visual exploration speed and errors during training.

Reading training (I2)

Each training trial was composed of the time-limited presentation of one single word in the centre of the screen. Patients were instructed to perceive each word as a whole before reading it aloud by intentionally shifting their gaze, as quickly as possible, from the screen’s centre to the beginning (in cases with

left-sided visual field loss) or to the end (in cases with right-sided visual

field loss) of each word.

During the course of training, the length of the presented words was systematically increased from 3- to 13-letter words. When a patient was able

to read at least 90% of the words of a given length correctly, presentation

time was reduced from ~1000 ms to 300–400 ms. The final training stage involved the randomized presentation of words of different lengths. This training protocol was adjusted to individual reading performance

and training progress. As above, an individual training session lasted ~45 min; it consisted of 10 practice units (30 trials each) and short or, if required, longer breaks between units. Training was completed when at least 90% of the responses were correct for any level of difficulty used. Patients required on average 12 training sessions, which were carried out within 2–3 weeks for each patient

An individual training session lasted ~45 min and consisted of 10 practice units (30 trials each).

The treatment was administered and supervised by the experimenter, who sat beside the patient to give

verbal feedback on reading or visual exploration speed and errors during training.

Length of follow-up:

12 weeks for each condition

Loss-to-follow-up:

Not reported

Incomplete outcome data:

Not reported

Visual field: Tübingen perimeter test

No data were presented.

Between-group difference in visual field sparing in degrees of visual angle from pre-to-post initial intervention:

T-test: p=0.26

Neglect: cancellation task

Mean number of errors (SD) at T3 versus T2:

I1:

Pre: 3.2 ±2.8

Post:0.5 ±0.7

Within-group difference: p=?

I2:

Pre: 2.9 ±2.5

Post: 3.1 ±2.6

Within-group difference: p=0.633

Vergelijking: interventie versus andere interventie

P=Stroke (34), tumour (2)

I1=visual scanning

12=reading

O=visual field (perimetry)

Wang, 2015

Type of study:

RCT

Setting and country:

China

Funding:

Government; University

Conflicts of interest:

None disclosed

Inclusion criteria:

-aged 30-80 years;

-dextromanual;

-right hemisphere infacrtion or haemorrhage;

-hemiplegia grade >=III level (Brunnstrom);

-after one month of onset.

Exclusion criteria:

-left hemianopia;

-aphasia or dementia;

-mental disorders;

-severe organ disease or unstable acute phase of the disease;

-deaf, dumb people;

-unwilling or unable to participate.

N total at baseline:

Intervention: 5

Control: 4

Important prognostic factors²:

age ± SD:

I: 64.6 ±10.55

C: 61.5 ±15.29

Sex:

I: 60% M

C: 75% M

Groups comparable at baseline?

yes

Conventional treatment plus Visual spatial attention

The conventional training was the same as in the other group.

Visual spatial attention training involved:

-optokinetic stimulation: In-house software on a displayed 150 stimuli squares on a 17-inch monitor with a black background. All simuli move in the leftward direction continuously at a speed of 7.5 */s. It lasts for 20 minutes. The patients were asked to pay attention to and watch this moving squares continuously for 20 minutes per day for 4 weeks.

-eye patching: standard spectacle frames containing non-corrective lenses with right half-field patches were given. Patching was done after any refraction problems were solved, using the patch and the patient’s corrective lenses. The patched glasses were worn 6 hours per day for 4 weeks. Patients were effectively forced to direct their gaze to their contralesional side.

Conventional treatment

Physical therapy (including neuromuscular facilitation, sit-to-stand training, balance training, treadmill training) and occupational therapy, including physical agents such as neuromuscular electrical stimulation. Participants trained twce a day, 6 days a week for 4 weeks.

Length of follow-up:

4 weeks

Loss-to-follow-up:

Intervention:

N (%)

Reasons (describe)

Control:

N (%)

Reasons (describe)

Incomplete outcome data:

Intervention:

N (%)

Reasons (describe)

Control:

N (%)

Reasons (describe)

Neglect: Chinese version of the Behavioural Inattention Test

Range 0277, higher scores indicating better performance.

I:

Pre: 77.20 ±20.91

Post: 132.00 ±30.07

Within-group mean difference: 54.8 ±11.03, T-test p<0.001

C:

Pre: 71.00 ±16.02

Post: 93.50 ±10.38

Within-group mean difference: 22.5 ±9.75, T-test p=0.02

Between-group difference: p=0.003

Vergelijking: interventie versus control

P=stroke

I=eye patching and optokinetic stimulation

C=usual care

O=neglect (BIT)

Study reference

Study characteristics

Patient characteristics

Intervention (I)

Comparison / control (C)

Follow-up

Outcome measures and effect size

Comments

Bowen, 2013

PS., study characteristics and results are extracted from the SR (unless stated otherwise)

SR and meta-analysis of RCTs

Literature search up to June 2012

A: Cherney, 2002

B: Cottam, 1987

C: Edmans, 2000

D: Fanthome, 1995

E: Ferreira, 2011*

F: Fong, 2007

G: Kalra, 1997

H: Kerkhoff, 2012a*

I: Luukkainen-Markkula, 2009*

J: Mizuno, 2011

K: Nys, 2008

L: Polanowska, 2009

M: Robertson, 1990

N: Robertson, 2002

O: Rossi, 1990

P: Rusconi, 2002

Q: Schroder, 2008*

R: Tsang, 2009

S: Turton, 2010

T: Weinberg, 1977

U: Welfringer, 2011

V: Wiart, 1997

W: Zeloni, 2002

*) study compared 2 cognitive interventions

Study design:

All RCTs

Setting and Country:

A: USA

B: USA

C: UK

D: UK

E: not reported

F: Hong Kong

G: UK

H: Germany

I: Finland

J: Japan

K: Netherlands

L: Poland

M: UK

N: UK

O: USA

P: Italy

Q: Germany

R: Hong Kong

S: UK

T: USA

U: Germany

V: France

W: Italy

Source of funding and conflicts of interest:

Not reported

Inclusion criteria SR:

-comparison between active treatment group that received one of various cognitive rehabilitation programs for neglect versus an alternative form of treatment or no-treatment control group.

-cognitive rehabilitation could include computerised therapy, prescription of aids and modification of environment as long as these were specific to neglect.

-trials that included participants with neglect following stroke, with stroke confirmed by neurological examination or brain scanning and neglect by neuropsychological assessment.

Exclusion criteria SR:

-drug treatments

-studies that included participants with head trauma, brain tumour, or any other brain damage unless a subgroup with stroke could be identified for which there were separate results or more than 75% of the sample had stroke.

-participants with general perceptual problems unless a subgroup with neglect could be identified.

23 studies included

Important patient characteristics at baseline:

N, mean age

AI: n=2, 69.5 ±23.3

AC: n=2, 62.0 ±5.7

BI: n=6, 66.2

BC: n=6, 71.3

CI: n=24, 69.2 ±11.4

CC: n=18, 66.6 ±14.5

DI: n=9, 66.3 ±10.7

DC: n=9, 71.1 ±7.6

EI+C n=10, -

FI1: n=20, 69.9 ±5.1

F2: n=20, 69.9 ±9.8

FC: n=20, 73.8 ±9.9

GI: n=24, 78 ±9

GC: n=76 ±10

HI: n=3?, 62.3

HC: n=3?, 56.3

II: n=6, 95.5 ±8.4

IC: n=6, 57.8 ±11.8

JI: n=20, 66.0 ±11.5

JC: n=18, 66.6 ±7.7

KI: n=10, 63.6 ±13.8

KC: n=6, 61.5 ±11.0

LI: n=20, 61.6 ±8.3

LC: n=20, 58.3 ±12.9

MI: n=17, 64.2 ±12.6

MC: n=13, 63.1 ±9.6

NI: n=19, 69.3 ±9

NC: n=21, 67 ±9.4

OI: n=18, 72.6

OC: n=21, 63.3

PI: n=12, 69.8

PC: n=8, 65.1

QI1: n=10, 68.4

QI2: n=10, 60.6

QC: n=10, 67.3

RI: n=17, 70.5 ±9.3

RC: n=17, 71.8 ±5.3

SI: n=17, 72 ±14

SC: n=19, 71 ±14

TI: n=14, 61.5 ±9.8

TC: n=11, 65.7 ±10.9

UI: n=15, 56.3 ±11.2

UC: n=15, 57.3 ±11.3

VI: n=11, 66

VC: n=11, 72

WI: n=4, 68.8

WC: n=4, 76.3

Sex:

AI: 100% Male

AC: 50%

BI+C: 58.3%

CI: 41.7%

CC: 44.4%

DI: 66.7%

DC: 66.7%

EI+C: -

F1+2+C: 56.7%

GI: 48.3%

GC: 39.1%

HI+C: -

II: 50%

IC: 33.3%

JI+C: -

KI: 70%

KC: 50%

LI: 55%

LC: 70%

MI: 45%

MC: 62.5%

NI: 68.4%

NC: 76.2%

OI: 55.5%

OC: 42.9%

PI: 41.7%

PC: 37.5%

QI1: 70%

QI2: 50%

QC: 60%

RI: 70.5%

RC: 52.9%

SI: 50%

SC: 57.9%

TI+C: -

UI+C: -

VI: 45.5%

VC: 45.5%

WI: 100%

WC: 50%

Groups comparable at baseline?

Not reported

A: Visual scanning, practising letter and word cancellation tasks using visual, verbal and tactile cues, 20 sessions

B: Five 30 min sessions per day of (phase 1) sessions of visual scanning a light board while stationary, (phase 2) visual scanning a light board while self-propelling, (phase 3) naming objects presented on both sides while self-propelling.

C: ToT approach: practicing perceptual task to treat the cause of the perceptual problem, 2.5 hrs per week for 6 weeks in addition to standard OT

D: feedback of eye movements (glasses with auditory signal), 2hrs and 40 min per week

E: visual scanning

F1: OT supervised ADL training and voluntary trunk rotation in ipsilateral reaching task (1 h/day, 5 days/week, 30 days)

F2: Same as F1 plus eye patching of ipsilesional field

G: Spatio-motor cuing based on attentional-motor integration and emphasis on restoration of function.

H: Optokinetic stimulation: repetitive leftward stimulation with active pursuit eye movements (moving stimuli), 20 sessions of 50 min, 5 sessions per week

I: Left arm activation training (20-30 hrs)

J: Pointing at targets without (30x), with (90x) and without (60x) prism-glasses (2x/day, 20 min, 5 d/week)

K: Prism adaptation: pointing at visual targets wearing prism goggles (30 min, 4 consecutive days)

L: Electrical somatosensory stimulation of the left hand (electrodes placed on the hand) plus visual scanning training using the RehaCom for computerised saccadic training and attention and concentration training and paper and pencil tasks to improve reading and writing; verbal and visual cues and instructions were given.

M: Computerized scanning and attention training (14 sessions of 75 min)

N: Wearing a limb activation device that emits a tone when no left-sided movement was made (45 min/week)

O: 15-diopter plastic press-on Fresnel prisms (cut in half circle, fitted in spectacle lenses with base toward affected field) worn during daytime activities in addition to standard rehabilitation including ADL training and visual perception training

P: visual scanning with cueing and feedback (reading sentences and stories, line drawing, assembling 3D cubes, matching cards, 40x1hr sessions)

Q1: visual exploration plus TENS (20x25-40 min)

Q2: visual exploration and optokinetic stimulation (20x25-40min)

R: eye patching glasses worn during conventional OT (30 min ADL, 30 min upper limb training, 4 weeks)

S: Prism adaptation training: repeated pointing to targets using unaffected hand while wearing prism glasses (10 sessions)

T: visual training: reading, writing, calculation (20 hours)

U: visuo-motor imagery therapy: mentally practicing positions and movements of the contralesional upper limb in a repetitive fashion (28-30 sessions) in addition to standardized rehabilitation including exploration training

V: Bon Saint Come method: wearing a thoracolumbar vest with metal pointer above the head, the participant points at targets om mobile panel, audible and luminous signals provide feedback (1 hr/day for 20 days), followed by 2-3 hours per day traditional rehabilitation

W: wearing goggles with eye patches for 1 week during waking hours

A: repetitive practice of functional task (oral reading) using visual, verbal and tactile cues, 20 sessions

B: Not described

C: FA approach: treat the symptom rather than the cause, 2.5 hrs per week for 6 weeks in addition to standard OT

D: No treatment

E: mental practice

F: Conventional OT (15 minutes of ADL and 45 min of upper extremity training)

G: conventional therapy concentrating on restoration of tone, movement pattern and motor activity before addressing skilled functional activity

H: visual scanning training on a computer screen (static stimuli), 20 sessions of 50 min, 5 sessions per week

I: Visual scanning training (10 hrs) combined with physiotherapy and OT/group therapy

J: Same task as in experimental group but wearing neutral glasses

K: Placebo: same as experimental group but wearing neutral goggles

L: Placebo: Same as in experimental group but without current intensity on the electrodes

M: recreational computing (games, quizzes and logical games, 11,4 hours)

N: Placebo: wearing the same limb activation device during perceptual training without tone (45 in/week).

O: No prism treatment in addition to standard rehabilitation including ADL training and visual perception training

P: Control: same activities without cueing and feedback (40x1hr sessions)

Q: Control: visual exploration using the ELEX apparatus (stimuli patterns presented on a screen)

R: Control: conventional OT without eye patches (30 min ADL, 30 min upper limb training, 4 weeks)

S: Sham treatment: same pointing activity while wearing plain glasses (10 sessions)

T: no visual training (but received OT as part of general rehabilitation program)

U: No visuo-motor imagery to standardized rehabilitation including exploration training

V: 2-3 hours traditional rehabilitation per day

W: No goggles

End-point of follow-up:

A: post-test

B: not reported

C: 6 weeks

D: 4 weeks

E: post-test and 3 months

F: 30 days (post-test) and 60 days

G: 12 weeks

H: 4 weeks

I: 3 weeks

J: 2 weeks and at hospital discharge

K: 1 month post treatment

L: 1 month (post-test)

M: 7 weeks

N: 12 weeks

O: 4 weeks

P: 2 months

Q: 4 weeks

R: 4 weeks

S: 2 weeks

T: 4 weeks

U: 3 weeks

V: 20 days

W: 1 week (post-test) and at 2 weeks

For how many participants were no complete outcome data available?

Not reported

Comparison 1: Cognitive rehabilitation versus any control

• immediate effects on ADL

Pooled SMD (random effects model): 0.23 (95% CI -0.02 to 0.48) favoring cognitive rehabilitation

Heterogeneity (I²): 24%

• Immediate effects on neglect

Pooled SMD (random effects model): 0.35 (95% CI 0.09 to 0.62) favoring cognitive rehabilitation

Heterogeneity (I²): 42%

• Delayed, follow-up effects on ADL

Pooled SMD (random effects model): 0.31 (95% CI -0.10 to 0.72) favoring cognitive rehabilitation

Heterogeneity (I²): 32%

• Delayed, follow-up effects on neglect

Pooled SMD (random effects model): 0.28 (95% CI -0.03 to 0.59) favoring cognitive rehabilitation

Heterogeneity (I²): 0%

Comparison 2: Cognitive rehabilitation versus another cognitive rehabilitation intervention

• immediate effects on ADL

Pooled SMD (random effects model): -0.28 (95% CI -1.15 to 0.59) favoring approach 2

Heterogeneity (I²): 0%

• delayed follow-up effects on ADL

Pooled SMD (random effects model): -0.48 (95% CI -1.54 to 0.58) favoring approach 2

Heterogeneity (I²): 31%

Comparison 3: visual scanning versus. another intervention

• immediate effects on neglect

Pooled SMD (random effects model): 0.09 (95% CI -0.71 to 0.89) favoring visual scanning

Heterogeneity (I²): 45%

• delayed follow-up effects on neglect

Pooled SMD (random effects model): -0.12 (95% CI -1.20 to 0.96) favoring approach 1

Heterogeneity (I²): 41%

Comparison 4: visual scanning versus. another visual scanning intervention

• immediate effects on neglect

Pooled SMD (random effects model): 0.95 (95% CI 0.43 to 1.47) favoring approach 1

Heterogeneity (I²): 45%

• delayed follow-up effects on neglect

Pooled SMD (random effects model): 1.13 (95% CI -0.33 to 2.60) favoring approach 1

Heterogeneity (I²): 65%

Comparison 3: prisms versus. any control*

• ADL

Pooled SMD (random effects model): 0.02 (95% CI -0.36 to 0.40) favoring prisms

Heterogeneity (I²): 0%

• neglect

Pooled SMD (random effects model): 0.29 (95% CI -0.16 to 0.74) favoring prisms

Heterogeneity (I²): 32%

Comparison 5: patching versus. any control*

• ADL

Pooled SMD (random effects model): 0.32 (95% CI -0.16 to 0.80) favoring prisms

Heterogeneity (I²): 0%

• neglect

Pooled SMD (random effects model): 0.47 (95% CI 0.01 to 0.93) favoring prisms

Heterogeneity (I²): 0%

Comparison 6: feedback or cueing versus. any control*

• ADL

Pooled SMD (random effects model): 0.40 (95% CI -0.48 to 1.28) favoring prisms

Heterogeneity (I²): 72%

• neglect

Pooled SMD (random effects model): 0.32 (95% CI -0.45 to 1.09) favoring prisms

Heterogeneity (I²): 70%

*) only immediate post-test results are presented. For delayed, follow-up effects, please refer to the original review.

Outcome: perimetry

The review did not look at perimetry. The following results were taken from the original publications:

A: not measured

B: could not access full text

C: not measured

D: could not access full text

E: not measured

F: not measured

G: not measured

H: not measured

I: not measured

J: not measured

K: not measured

L: not measured

M: could not access full text

N: not measured

O: could not access full text

P: could not access full text

Q: not measured

R: not measured

S: not measured

T: could not access full text

U: not measured

V: not measured

W: not measured

Risk of bias

-Allocation concealment (6 high risk, 6 unclear)

-Blinding outcome assessors (3 high risk, 6 unclear)

-Incomplete outcome data (4 high risk, 4 unclear)

-Free of systematic differences between groups at baseline (7 high risk, 7 unclear)

-Adequate adjustment for differences at baseline (5 high risk, 4 unclear)

-Potential bias due to blinding of participants, blinding of therapists and violation of intention to treat were not assessed.

Author’s conclusion

The effectiveness of cognitive rehabilitation interventions for reducing the disabling effects of neglect and increasing independence

remains unproven. As a consequence, no rehabilitation approach can be supported or refuted based on current evidence from RCTs.

However, there is some very limited evidence that cognitive rehabilitation may have an immediate beneficial effect on tests of neglect. This emerging evidence justifies further clinical trials of cognitive rehabilitation for neglect. However, future studies need to have appropriate high quality methodological design and reporting, to examine persisting effects of treatment and to include an attention

control comparator.

Study reference

(first author, publication year)

Describe method of randomisation¹

Bias due to inadequate concealment of allocation?²

(unlikely/likely/unclear)

Bias due to inadequate blinding of participants to treatment allocation?³

(unlikely/likely/unclear)

Bias due to inadequate blinding of care providers to treatment allocation?³

(unlikely/likely/unclear)

Bias due to inadequate blinding of outcome assessors to treatment allocation?³

(unlikely/likely/unclear)

Bias due to selective outcome reporting on basis of the results?⁴

(unlikely/likely/unclear)

Bias due to loss to follow-up?⁵

(unlikely/likely/unclear)

Bias due to violation of

intention to treat analysis?⁶

(unlikely/likely/unclear)

Aparicio-Lopez, 2016

Once the informed consent of the participants had been obtained they were allotted a study number. These numbers corresponded to the order in which the patients entered the study. A simple randomization procedure was performed according to a computer-generated random number table based on a uniform distribution (0, 1). The application of such a simple randomization procedure yielded 15 patients in the ST group and 13 in the CT group. The research assistant who generated the allocation scheme was not clinically involved in

the study (neither in the assessment nor in the administration of

treatment to the patients).

Unlikely

Unclear

Likely

likely

Unlikely

Unlikely

Unclear

Kerkhoff, 2014

Finally, 24 patients with acute visuospatial neglect, as assessed by the 2 screening tests,

were randomly allocated to SPT or VST (by a person not

involved in the study who drew lots from a sealed envelope).

Unlikely

Unclear

Unclear

Unlikely

Unlikely

Unlikely

Unclear

Keller, 2010

Using a random number algorithm,

patients were assigned to either audiovisual stimulation training or visual stimulation training. To achieve a nearly balanced number of patients with quadrantanopia in

each group, the assignment of quadrantanopic patients between

treatment groups was alternated, resulting in 3 patients in the AVT group and 4 patients in the VT group.

Partly unlikely (hemianopic patients), partly likely (quadrantanopic patients)

Unlikely

Unclear

Unclear

Unlikely

Unclear

Unclear

Modden, 2012

Patients were randomly assigned to receive either CT, RT, or OT. Randomization by throwing dice and allocation took place before starting with the initial assessment of neuropsychological tests.

Unlikely

Unclear

Unclear

Likely

Unlikely

Unlikely

Unclear

Crotty, 2018

Those who met the criteria and agreed

to participate provided the research therapist written informed consent and, after baseline assessment,

were block randomized in groups of eight. Randomization was performed by a statistician from the

pharmacy department who was not part of the study, and generated the random sequence using the random number generator in Microsoft Excel, creating

sequentially numbered, sealed envelopes containing group allocation for participants.

Unlikely

Likely

Likely

Unlikely

Unlikely

Likely

Likely

Rowe, 2017

Participants were individually randomized to one of three treatment groups using a secure (24-hour) web-based randomization programme.

Randomization lists were generated using block randomization stratified by centre and degree of hemianopia (partial or complete) with treatment allocation ratio of 1:1:1. The local PI (orthoptist) obtained

the treatment allocation and subsequently assigned the participant to the treatment arm.

Unlikely

Unclear

Unclear

Unclear

Unlikely

Likely

Likely

Elshout, 2016

To prevent selection bias, prior to the inclusion of the first patient, a training scheme for all patients was created using Matlab (Mathworks Inc.). For each cohort of 10 patients, numbers (J01–J40), the training stimuli (Flow or Point), and the order of training rounds (Test (defect) round first or Control (intact) round first) were randomly assigned to a patient number (Table S1 in Supplementary Material). The patients were included in order of registration on our website. Thus, the assignment of a patient to a patient number (with corresponding training scheme: stimulus type and the training order) was determined prior to the first inclusion, completely random and not based on selection.

Unlikely

Unlikely

Likely

Unclear

Unlikely

Likely

unclear

De Haan, 2015

Allocation to the groups occurred by the method of minimization (37).

Differences between the groups were minimized regarding gender, side of field defect (left or right), size of field defect (hemianopia versus. quadrantanopia),

age (younger versus. older than 55), and time since onset (shorter versus. longer than 12 months). Because time since onset was assumed less important than the other variables, this variable was weighted less heavily (0.5) than the others (1.0). Upon inclusion of a new participant, author GH entered the characteristics of the patient into the randomization software that contained the characteristics of the previously included patients, which resulted in allocation to

the training group or the waiting list control group.

Unlikely

Unclear

Unclear

Unlikely

Unlikely

Likely

Likely

Schuett, 2012

Patients were randomly allocated into two treatment groups.

Unclear

Unclear

Likely

Unclear

Unlikely

Unclear

Unclear

Wang, 2015

The 9 cases were randomly divided into conventional treatment group + visual spatial attention and conventional treatment group.

Unclear

Unclear

Unclear

Unclear

Unlikely

Unclear

unclear

Study

First author, year

Appropriate and clearly focused question?¹

Yes/no/unclear

Comprehensive and systematic literature search?²

Yes/no/unclear

Description of included and excluded studies?³

Yes/no/unclear

Description of relevant characteristics of included studies?⁴

Yes/no/unclear

Appropriate adjustment for potential confounders in observational studies?⁵

Yes/no/unclear/notapplicable

Assessment of scientific quality of included studies?⁶

Yes/no/unclear

Enough similarities between studies to make combining them reasonable?⁷

Yes/no/unclear

Potential risk of publication bias taken into account?⁸

Yes/no/unclear

Potential conflicts of interest reported?⁹

Yes/no/unclear

Bowen, 2013

Yes

Yes

Yes

Yes

n/a

Yes

Yes

No

No