Carter 1983.

Study characteristics
Methods	Design: parallel RCT; 1988: post‐hoc analysis of data from Carter 1983 Duration of trial: not reported Unit of randomisation: inpatient adults with a clinically defined stroke and cognitive impairment Recruitment and allocation: 33 people with acute stroke were randomised to 2 groups, IG (n = 16) and CG (n = 17)
Participants	Setting: metropolitan community hospital Country: USA Sample size: 33 adults (48% men); IG: 16; CG: 17; n = 28 analysed for ADL, n = 25 analysed for time judgement Exclusion criteria: people with tumours, extensive bilateral damage, or prior brain damage Age: mean: IG: 70.5 (SD 11.4) years; CG: 73.4 (SD 9.2) years Types of stroke: neurological Severity Scores mean (of 60): IG: 29.4 (SD 3.9); CG: 28.5 (SD 6.4) Site of lesion: IG: left hemisphere 9 (56%), right 7 (44%); CG: left hemisphere 9 (53%), right 8 (47%) Days from admittance to stroke programme to pretest: mean: IG: 4.8 (SD 1.6) days; CG: 4.6 (SD 2.6) days
Interventions	Intervention group Brief name: cognitive skills remediation training Recipients: inpatient adults with cognitive impairments after acute stroke Why: to provide formal cognitive remediation training to acute patients within 1 week poststroke that included "continuous reinforcement, immediate feedback, cuing, gradually increasing … difficulty level and stressing the importance of the skills being taught to activities of daily living" (Carter 1983). What (materials): Thinking Skills Workbook (Carter 1980) for pen‐and‐paper tasks requiring: visual scanning, visual‐spatial or time judgement; early versions cited in papers, latest edition (Languirand 2014) provided pen‐and‐paper pre‐ and post‐tests and tasks covering various areas of cognition including paying attention and reading, concentrating on detail, listening, scheduling and time management, memory in everyday living, sorting and classifying, sequencing and logic, verbal skills, maths skills. What (procedures): based on pretest of the 3 main areas of interest, visual scanning, visual‐spatial, or time judgement, trained research assistants provided 1‐to‐1 training sessions in any of these areas where pretest performance was < 80%, in additional to usual rehabilitation. Who provided: trained research assistants provided the intervention in the study; Languirand 2014 workbook stated that it was for use by professional rehabilitation staff, paraprofessionals, or family members. How: face‐to‐face and individually Where: stroke rehabilitation unit; Languirand 2014 workbook recommended a quiet private room with minimal distractions. When and how much: 30–40 minutes, 3 times per week for a mean of 3–4 weeks (i.e. up to 8 hours in total); the testing and training took place between 9.00 a.m. and 11.45 a.m. before or after other stroke programme activities; Languirand 2014 workbook recommended session length of 25–35 minutes, twice per week for 4–6 weeks Tailoring: not reported except that training was given only in areas where the participant scored < 80%. Modifications: none reported How well (planned): only that the research assistants were trained in activities from the workbook. How well (actual): not reported Comparator group Brief name: rehabilitation as usual Recipients: inpatient adults with cognitive impairments after acute stroke Why: to provide rehabilitation services poststroke What: routine stroke programme activities, including social work consultations, occupational therapy, physical therapy, speech therapy, family visits, and interaction with rehabilitation nursing staff Who: occupational therapists, physical therapists, speech therapists, social workers, and rehabilitation nursing staff How: face‐to‐face Where: stroke rehabilitation unit When and how much: mean length of stroke programme was 3–4 weeks Tailoring: not reported Modifications: not reported How well (planned): not reported How well (actual): not reported
Outcomes	Primary Barthel Index Secondary Time judgement test: participants were asked to estimate a 1‐minute time period Other Visual scanning, visual‐spatial tests Methods of data collection: attempts were made for blinded pre‐ and postassessment by trained trainers and testers not involved in stroke programme. Data collection time points: before and after training (baseline and 3–4 weeks after baseline)
Notes	Funding: yes (research grant) Conflict of interest: not reported Published trial protocol: no Trial registration: no Ethics approval: none reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Description provided was "randomly assigned".
Allocation concealment (selection bias)	Unclear risk	Not reported.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Quote: "The stroke program staff was not informed of group assignment for the patients in this project, or of the experimental nature of the project … However, because of the physical layout of the stroke unit, at times it was possible for the tester to see which patients were given training by the other assistant".
Blinding of outcome assessment (detection bias) All outcomes	High risk	Incomplete blinding. Quote: "For the major part of the study, an experimental blind testing procedure was used … However … at times it was possible for the tester to see which patients were given training by the other assistant".
Incomplete outcome data (attrition bias) All outcomes	Low risk	Overall improvement scores are provided for the IG (n = 16) and CG (n = 17), suggesting that there was no attrition. In the secondary analysis, data from 28/33 participants was available for the Barthel Index with attrition equal across groups.
Selective reporting (reporting bias)	High risk	Post‐hoc analysis of ADLs reported in Carter 1988 but not referred to in Carter 1983; Carter 1988 also reported auditory attention, digit span, verbal memory, abstract reasoning, and verbal comprehension (that are not reported in Carter 1983).
Other bias	Low risk	No other identifiable bias.