日本医療・統計研究会雑誌日本語原著研究

randMS No.1 原著 研究

要旨
近年の臨床治験に関連する論文は,Consolidated Standards of Reporting Trials (CONSORT)声明に準拠したランダム化実施の明記がなければ,査読前の段階で受理されることが難しい.また,ランダム化のソフトウエア(以後ソフト)は高額であり,一般臨床医が容易に利用できる環境にない.
自動でランダム割付を実施する本ソフト(randMS:ランダ・エムエス)は,Filemaker®社のFilemaker pro advanced®で作成した.randMSの特徴は,事前・事後の層別ができ,静的割付(ランダム置換ブロック法)・動的割付(最小化法)の選択ができ,ランダム化ができるように設計した.
一般的なランダム化比較試験の概説とともに,randMSのランダム化の手順を紹介し,無料で公開した.臨床治験に興味ある医師のリサーチマインドをサポートすることが第一の目的とし,randMSの修正・改善に加え,利用者間の情報の共有化することで,将来的には大学病院医療情報ネットワーク(UMIN)に推奨されるソフトを目指すことを第二の目的とした.
Abstract
Recent reports of clinical trials that do not include a statement of conformance with Consolidated Standards of Reporting Trials (CONSORT) are generally returned to the authors before going through the review stage. Because software for randomization is generally too expensive, most clinicians cannot easily adopt such software for this purpose. The program randMS (randa-em-ess), which can perform automatic randomized allocation, was developed in the software Filemaker Pro Advanced®. The characteristics of randMS enable performing stratification both before and after clinical trials. Furthermore, randMS can select from a fixed-allocation design such as the use of randomly permuted blocks or a dynamic-allocation design such as a minimization method. This report provides an outline for a randomized controlled trial, and introduces the steps used for randomization with the software, which is made freely available. The target audience of this report is general clinicians who are interested in clinical trials to support basic research. Furthermore, randMS was modified and developed to enable sharing information between users, with the goal of becoming the recommended software for randomization in University Hospital Medical Information Network (UMIN) in the future.
pdfNo.1

おなまえ 
タイトル 

投稿者パスワード 編集・削除時に必要です