2018 Fiscal Year Final Research Report
Alternative methods of evaluating treatment effect in case of correlated competing risks in clinical trials
Project/Area Number |
16K12400
|
Research Category |
Grant-in-Aid for Challenging Exploratory Research
|
Allocation Type | Multi-year Fund |
Research Field |
Statistical science
|
Research Institution | Niigata University |
Principal Investigator |
Akazawa Kohei 新潟大学, 医歯学総合病院, 教授 (10175771)
|
Project Period (FY) |
2016-04-01 – 2019-03-31
|
Keywords | 生存時間解析 / 競合リスク / 無再発期間 / ログランク検定 / 比例ハザードモデル |
Outline of Final Research Achievements |
The results of this study is following: (1) We estimated the frequencies of cause of death (cancer-related death, progressive failure of physical condition, other causes of death), and investigated a relationship between progression-free survival and overall survival; (2) The properties of logrank test such as size of test and power were evaluated by statistical simulation, in case of the presence of competing risks in cause-specific survival analysis; (3) We found out the statistical problems on applying logrank test and Cox proportional hazards model for multiple endpoints.
|
Free Research Field |
医療統計学
|
Academic Significance and Societal Importance of the Research Achievements |
第Ⅱ相、第Ⅲ相臨床試験を計画する際に、評価尺度の選定はその試験デザインをほぼ決定する。代理評価尺度を主要評価尺度として採用する場合に、真の評価尺度との関連の度合いを定量的に把握できれば、不必要な代理評価尺度を選定せずに済み、臨床試験の質を向上させる可能性がある。逆に、代理評価尺度が真の評価尺度と強い関連を持つのであれば、副次評価尺度として生存期間の長期追跡調査は不要となり、臨床試験期間の短縮、試験費用の軽減、新薬の早期の承認につながる。
|