Duncan 1998

Methods	RCT Participants randomly assigned to control or intervention group using a random list generated by group assignments. Randomisation completed in blocks of 10. Random list generated prior to the beginning of the study. Only a laboratory technician who had no input into participant selection or recruitment was aware of group assignment
Participants	20 participants selected from local participating hospitals and Kansas City Stroke Registry. To be included on this registry participant had to have a stroke defined by WHO definition Inclusion criteria: 30 to 90 days after stroke; minimal or moderately impaired sensorimotor function (Fugl‐Meyer 40 to 90, Oprington Prognostic Scale score 2.0 to 5.2); ambulatory with supervision and/or assistive device; living at home; living within 50 miles of the University of Kansas Medical Center; no medical condition that interfered with outcome assessments or limited participation in submaximal exercise programme; MMSE > 18; and no receptive aphasia that interfered with ability to follow a 3‐step command
Interventions	Group 1 (10 participants): usual care. Usual care as prescribed by their physicians. Reserach assistant visited every 2 weeks to assess the participants exercise and activity level. Clinicans completed an intervention log to capture type of exercises and frequency and duration of therapy visits during treatment or in a home exercise programme. The therapy programmes received by the control group varied in intensity, frequency and duration Group 2 (10 participants): home therapy programme. This involved an exercise programme designed to improve strength, balance and endurance and to encourage more use of the affected extremity. No other physical or occupational therapy was provided. The programme was a home‐based exercise programme provided by a physical therapist. The study principal investigator (physiotherapist) and co‐investigator (occupational therapist) observed at least 1 therapy session for each participant to ensure standard application of interventions. Exercise sessions were divided into the following 4 blocks (preceded by a 10‐minute warm‐up session of stretching and flexibility exercises) Assistive and resistive exercises using PNF patterns or theraband exercises to the major muscle groups of the upper and lower extremities Balance exercises Encouraged to use the affected upper extremity in functional activities Progressive walking programme or progressive exercise on a bicycle ergometer The programme included 3 visits per week for 8 weeks, and the patients were instructed to continue the exercise programme for an additional 4 weeks. Each session lasted approximately 1.5 hours
Outcomes	Primary outcome 1: performance in ADL: Barthel Index (0 to 100) Primary outcome 2: functional movement: Jebsen Test of Hand Function (dexterity measure). We could not include data for this outcome in the data analysis as total scores and SD were not reported Secondary outcome 1: performance in extended ADL: Lawton Instrumental ADL Secondary outcome 2: (motor impairment) Fugl‐Meyer Upper Extremity Scale (0 to 66) Oprington Prognositc Scale, Fugl‐Meyer Lower Extremity Scale (0 to 34), Medical Outcomes study‐ 36 Health Status Measurement, 10 metre walk, 6 minute walk and Berg Balance Scale were also reported, but are not relevant to this review Outcome measures completed at the end of intervention period only
Notes	SDs are not included in the paper. However, we were able to calculate the SDs from data gained from the study authors. Data gained from study authors was also used to enter mean values for Barthel Index. This data gained from personal communication with the author differs from those presented in the published paper
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	"A random list was generated by group assignments"
Allocation concealment (selection bias)	Low risk	"Only a laboratory technician who had no input into subject selection or recruitment was aware of group assignment"
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Comment: no information provided
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Comment: no dropouts for any of the reported outcomes
Selective reporting (reporting bias)	Low risk	One outcome not reported (10 metre walk test), however all other pre‐specified outcomes were reported