脳卒中後上肢麻痺における回復期の傾向スコアデータプールの構築

Constructing a propensity score data pool for the subacute in upper limb paralysis after stroke

要旨：エビデンスレベルの高い介入研究を実施するためには，対照群を設定する必要があるが倫理的に容易ではない．本研究では，一般臨床のデータを後方視的に抽出することにより，214名の脳卒中後上肢麻痺における回復期の傾向スコア算出のためのデータプールを構築した．データの使用例として，過去に著者らが実施した，対照群を用いない前後比較試験のデータを介入群とした傾向スコアマッチングを行った結果についても併せて報告する．本研究のデータを多くの研究者や臨床家が対照群として使用することにより，臨床研究の質向上に貢献できると考える．

　To conduct high-level intervention studies, it is necessary to set up a control group, which presents ethical difficulties. Therefore, in this study, we decided to construct a data pool for calculating the propensity score of the recovery period in 214 upper limb paralysis after stroke by retrospectively extracting data that was conducted as an evaluation in general clinical practice. We also report the results of propensity score matching in which data from previous and subsequent comparative studies that did not use control groups performed by the author as an intervention group was used as an example of data usage. By sharing the data of this study with many researchers and clinicians and using it as a control group, we can improve the accuracy of clinical studies.