Risk-based monitoring of clinical data

A computing device for generating an audit trail for monitoring clinical data, the computing device comprising:

• interface circuitry configured to exchange signals over one or more data paths of the computing device;
• processing circuitry communicatively coupled to the interface circuitry and configured to:
• receive clinical data comprising patient data and investigator site data via the interface circuitry, wherein the investigator site data comprises safety metrics;
• accept and reject respective portions of the clinical data by applying a risk-based monitoring policy comprising one or more policy rules to the clinical data, wherein applying the risk-based monitoring policy comprises applying, to the clinical data, a machine-learning algorithm, wherein applying the machine-learning algorithm comprises:
• learning, over time, patterns in the patient data and investigator site data of the clinical data;
• receiving, over time, input from users regarding risk indicators in the learned patterns;
• and
• maintaining, over time, a knowledge library based on the learned patterns and received input, wherein the accepting and rejecting of the portions of the clinical data is based on the maintained knowledge library;
• generate an audit trail for the clinical data based on applying the risk-based monitoring policy to the clinical data, the audit trail comprising the clinical data and corresponding audit data;
• wherein the audit data comprises:
• identification of the accepted and rejected portions of the clinical data;
• identification of one or more policy rules of the risk-based monitoring policy used in the accepting and rejecting of the respective portions of the clinical data;
• one or more timestamps corresponding to when each of the accepting and rejecting of the respective portions took place.

A non-transitory computer readable medium storing a computer program product for controlling a programmable computing device in a communication network, the computer program product comprising software instructions that, when run on the programmable computing device, cause the programmable computing device to:

• receive clinical data comprising patient data and investigator site data via interface circuitry, wherein the investigator site data comprises safety metrics;
• accept and reject respective portions of the clinical data by applying a risk-based monitoring policy comprising one or more policy rules to the clinical data, wherein applying the risk-based monitoring policy comprises applying, to the clinical data, a machine-learning algorithm, wherein applying the machine-learning algorithm comprises:
• learning, over time, patterns in the patient data and investigator site data of the clinical data;
• receiving, over time, input from users regarding risk indicators in the learned patterns;
• and
• maintaining, over time, a knowledge library based on the learned patterns and received input, wherein the accepting and rejecting of the portions of the clinical data is based on the maintained knowledge library;
• generate an audit trail for the clinical data based on applying the risk-based monitoring policy to the clinical data, the audit trail comprising the clinical data and corresponding audit data, the audit data comprising:
• identification of the accepted and rejected portions of the clinical data;
• identification of one or more policy rules of the risk-based monitoring policy used in the accepting and rejecting of the respective portions of the clinical data;
• one or more timestamps corresponding to when each of the accepting and rejecting of the respective portions took place.