Abstract
Much of the research in upper limb dysfunction has targeted the chronic phase. However, studies with robust results in the acute phase are lacking. Due to the difficulty in conducting comparative studies during the acute phase we periodically extracted the evaluation data of stroke-related upper limb paralysis and constructed a data pool from the calculations of the propensity scores. This data pool enabled the creation of a control group in the acute phase that is normally difficult to establish. The data pool also granted the ability of many institutions to share the data, resulting in verification of new interventions. In addition, as an example of how this data could be utilized, we report the results of propensity score matching using the intervention group during the current period.