| 摘要 |
A system for developing and implementing empirically derived algorithms to generate decision rules to determine participant noncompliance and fraud with research protocols in clinical trials allows for the identification of complex patterns of variables that detect or predict participant noncompliance and fraud with research protocol, including performance and enrollment goals, in the clinical trial. The data may be used to overall predict the performance of any participant in a clinical trial, allowing selection of participants that tend to produce useful, high-quality results. The present invention can also be used to monitor participant compliance with the research protocol and goals to determine preferred actions to be performed. Optionally, the invention may provide a spectrum of noncompliance, from minor noncompliance needing only corrective feedback, to significant noncompliance requiring participant removal from the clinical trial or from future clinical trials. The algorithms and decision rules can also be domain-specific, such as detecting non-compliance or fraud among subjects in a cardiovascular drug trial, or demographically specific, such as taking into account gender, age or location, which provides for algorithms and decision rules to be optimized for the specific sample of participants being studied.
|
