Lang 2016c.

Study characteristics
Methods	Parallel RCT Participants who met the selection criteria were randomised into 4 groups, each received different number of repetitions
Participants	85 participants in 4 study groups 3.2k reps group: n = 21; mean age 59.9 years; 7 females; 12.0 months poststroke 6.4k reps group: n = 22; mean age 62.1 years; 5 females; 13.0 months poststroke 9.6k reps group: n = 21; mean age 60.0 years; 10 females; 13.0 months poststroke Individualised maximum group: n = 21; mean age 60.9 years; 8 females; 11.5 months poststroke Outpatient setting
Interventions	Intervention was supervised, massed practice of functional upper limb daily tasks e.g. reaching, grasping, moving/manipulating, and releasing object. The participants in the 4 groups were encouraged to perform a certain number of repetitions of exercise, dependent on their group allocation Number of reps per session were: 100, 200, or 300. The individualised maximum group aimed for 300 reps per session, but to continue receiving therapy until certain criteria had been met
Outcomes	ARAT (primary outcome measurement) SIS (hand and ADL subscales) COPM 7‐point Likert scale to evaluate if the participant though they had changed, and if that change was meaningful Measures taken at baseline (prior to randomisation, postintervention, and then 2 months later
Notes	Lang 2016c refers to the  following pair‐wise comparison: 3.2k rep (n = 6) vs individualised maximum group (n = 18) This study specified repetitions of exercise, as opposed to time spent, but also report the number of minutes of 'active practice' undertaken in each group Conflict of interest: none Funding: NIH R01 HD068290