Bo 2019.

Study characteristics
Methods	Design: parallel RCT with 4 groups (note: data only for 2 groups included in this review: CG and CT groups) Duration of trial: about 11 months (first recruitment 10 February 2017 to 22 January 2018 (last follow‐up)) Unit of randomisation: outpatients < 6 months poststroke with vascular cognitive impairment (aged > 18 years) Recruitment and allocation: of 260 people with stroke assessed for eligibility, 225 people were eligible and randomised to 1 of 4 groups: PE + CT; PE; CT; and CG; as noted above, we included only data from the CT (n = 57) and CG (n = 57)
Participants	Setting: rehabilitation centre Country: China Sample size: 114 allocated, 92 adults analysed; CT: 45 (53.33% men); CG: 47 (57.44% men) Inclusion criteria: aged > 18 years; medically stable; < 6 months poststroke; able to understand and follow verbal instructions; without severe somatic diseases or mental disorders, including anxiety and depression; without visual or auditory disturbances in recent months; met the diagnostic criteria for vascular cognitive impairment Exclusion criteria: motor deficits, non‐stroke‐related neurological impairments, clinically determined as unsafe for physical activity Age: mean: CT: 67.51 (SD 2.24) years; CG: 64.36 (SD 2.31) years Time since onset of stroke: not reported; however, inclusion criterion of < 6 months poststroke Types of stroke: not reported Site of lesion: not reported
Interventions	Intervention group Brief name: CT (CO) Recipients: outpatients < 6 months poststroke with vascular cognitive impairment (aged > 18 years) Why: to provide an effective and safe alternative to established drugs to decrease cognitive impairments in people with stroke in the form of a non‐pharmacological intervention of CT. What (materials): COGPACK programme, developed for neurorehabilitation (Marker Software, www.markersoftware.com) delivered on up to 20 tablet computers with touch screens "to avoid training difficulties in computer novices"; 12 of possible 64 exercises were selected, including 4 tasks of memory ('memory for route', 'memory for signs', 'memory for pattern', 'memory for scene'), 4 tasks of execution ('mental arithmetic', 'logical block', 'shortest way', 'continue a series'), and 4 attention and speed tasks ('scanning', 'catch', 'steer', 'assembly line'). What (procedures): supervised CT in groups using tablet computers Who: experienced therapists with exercise physiology or clinical psychology backgrounds provided the interventions for the 4 groups. However, the CT group used the COGPACK programme, which is a commercially available programme and which, according to the website, can be used in occupational therapy as a "concentration, performance and motivation aid". How: supervised group (up to 20) Where: in the rehabilitation centre, further details not provided When and how much: 60 minutes, 3 times weekly, for 12 weeks (36 hours in total) Tailoring: group training was supervised, so presumably individual and group assistance was provided as needed. Modifications: none reported How well (planned): not reported How well (actual): not reported Comparator group Brief name: usual care and watched 45‐minute video documentaries 3 times per week, for 12 weeks Recipients: outpatients < 6 months poststroke with vascular cognitive impairment (aged > 18 years) Why: to provide usual rehabilitation services to patients poststroke and an equivalent dose through the documentary watching. What: usual care and video documentaries Who provided: not described, presumably usual care rehabilitation staff How: face‐to‐face Where: rehabilitation centre When and how much: length of usual care plus 45‐minute video documentaries 3 times per week, for 12 weeks Tailoring: not reported Modifications: not reported How well (planned): not reported How well (actual): not reported
Outcomes	Primary None Secondary TMT‐B Stroop Colour‐Word Forward Digit Span Mental Rotation Test Methods of data collection: 4 research assistants blinded to group allocation performed a battery of standardised measurements. Data collection time points: baseline, 12 weeks (postintervention), and 6 months after training ended
Notes	Funding: none. Authors declared they received no financial support for the research, authorship, or publication of the study. Conflict of interest: the authors declared no potential conflicts of interest with respect to the research, authorship, or publication of the study. Published trial protocol: none reported Trial registration: ISRCTN16009172 Ethics approval: yes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "used the on‐line Research Randomizer to generate allocation sequence".
Allocation concealment (selection bias)	Low risk	Quote: "an independent research assistant who was not involved in the study held the random lists of number" and "contacted the participants".
Blinding of participants and personnel (performance bias) All outcomes	High risk	Incomplete blinding. Single‐blind (investigator‐blinded, not participant blinded); it is not clear if the "investigator" is the researcher or the therapist, may have been only the outcome assessors (research assistants). Therefore, it may have been possible for therapists to discern which therapy a participant was receiving and for the participants to know which group they were allocated to.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote: "Four research assistants performed a battery of standardized measurements … The research assistants were blind to group allocation".
Incomplete outcome data (attrition bias) All outcomes	High risk	Loss to follow‐up and the reasons for losses were reported and these were balanced between groups. However, 12 (21%) missing from CT group and 10 (17.5%) from CG.
Selective reporting (reporting bias)	Low risk	Checked against protocol and all outcomes assessed and reported as planned.
Other bias	Low risk	No other identifiable bias.