Detection of outlying centers and influence diagnostics for the analysis of multicenter clinical trials

DOI Web Site Web Site 29 References Open Access
  • Nakamura Rie
    KOSÉ Corporation Research Laboratories Department of Statistical Science, School of Multidisciplinary Sciences, The Graduate University of Advanced Studies
  • Noma Hisashi
    Department of Data Science, The Institute of Statistical Mathematics

Bibliographic Information

Other Title
  • 多施設共同臨床試験における極端なプロファイルを持つ施設の検出と影響力診断の方法
  • タシセツ キョウドウ リンショウ シケン ニ オケル キョクタン ナ プロファイル オ モツ シセツ ノ ケンシュツ ト エイキョウリョク シンダン ノ ホウホウ

Search this article

Abstract

<p>In multicenter clinical trials, the assessment for heterogeneity of various relevant factors across participating centers is a relevant issue because it can cause inconsistency of the treatment effects. Especially, outlying centers with extreme profiles can influence the overall conclusions of these trials. In this article, we propose quantitative methods to detect the outlying centers and to assess their influences in multicenter clinical trials. We proposed four effective methods based on (1) a studentized residual obtained by a leave-one-out analysis, (2) a model-based significance test to detect an outlying trial using a mean-shifted model, (3) a relative change measure for the variance estimate of the overall treatment effect estimator, and (4) a relative change measure for the heterogeneity variance estimate in a random-effects model. In addition, we provide parametric bootstrap algorithms to assess the statistical variability of their influential measures. We also demonstrate the practical effectiveness of these proposed methods via applications to two clinical trials for benign prostatic hyperplasia and cardiovascular heart disease.</p>

Journal

References(29)*help

See more

Related Projects

See more

Keywords

Details 詳細情報について

Report a problem

Back to top