2022 Fiscal Year Final Research Report
Research on randomization methods and comparative methods of screening tests in small clinical trials
Project/Area Number |
18K17325
|
Research Category |
Grant-in-Aid for Early-Career Scientists
|
Allocation Type | Multi-year Fund |
Review Section |
Basic Section 58010:Medical management and medical sociology-related
|
Research Institution | Hyogo Medical University (2021-2022) Osaka City University (2018-2020) |
Principal Investigator |
|
Project Period (FY) |
2018-04-01 – 2023-03-31
|
Keywords | ランダム化 / 陽性的中率 / 陰性的中率 |
Outline of Final Research Achievements |
The purpose of this study was to propose guidance for the selection of randomization methods and an to propose analysis methods for comparison of the performance of screening tests in small-size clinical trials. Regarding the choice of randomization method, we showed that randomization methods that consider prognostic factors perform better in small clinical trials, and that there is little difference in performance among randomization methods that take prognostic factors into account. For analysis methods to compare the performance of screening tests, we developed exact tests and non-inferiority tests for positive and negative predictive values, as well as a simultaneous test procedure to demonstrate the superiority of either positive or negative predictive value, the non-inferiority of the other.
|
Free Research Field |
生物統計学
|
Academic Significance and Societal Importance of the Research Achievements |
小規模臨床試験に対しては,単純なランダム化では群間で偏りが生じる可能性が高いため,動的割り付けなど複雑な手法が選択される傾向にある.小規模臨床試験に着目してアウトカムの尺度別に割付手法の比較を行った本研究の成果は,小規模臨床試験における割付方法の選択の一助になると考えられる.また,スクリーニング検査法の解析手法に関しては,陽性的中率および陰性的中率の検定手法として,小規模臨床試験にも適用可能な正確検定のほかに,非劣性検定や同時検定手順についても開発した.これらの検定手法は,小規模臨床試験を含め,今後の陽性的中率および陰性的中率に関する研究に役立つと考えられる.
|