Considerations for collecting clinical data and performing Data Management

Considerations for collecting clinical data and performing Data Management at the launch of a Computer-Assisted Detection Devices (CADe) study, Applied to Radiology Images and Radiology Device Data in – Premarket Notification 510k Submissions

CADe devices are computerized systems that incorporate pattern recognition and data analysis capabilities from patient radiological data, intended to identify, mark, highlight, or in any other manner direct attention to portions of an image, or aspects of radiology device data, that may reveal abnormalities during interpretation of patient radiology images or patient radiology device data by the intended user (i.e., a physician or other health care professional).

Study design:

The clinical performance assessment of CADe device is typically performed by utilizing a multiple reader, multiple case (MRMC) study design.  A set of clinical readers (i.e., clinicians evaluating the radiological images or data in the MRMC study) evaluate image data under multiple reading conditions or modalities (e.g., readers unaided by CADe versus readers aided by CADe). The MRMC design can be “fully-crossed” whereby all readers independently read all of the cases.

Electronic Data Capture (EDC) should support your study with the relevant aspects:   

• Randomization Module, there are few reading scenarios that may be part of a CADe clinical evaluation. Readers, cases, and reading scenarios should be randomized to reduce bias in performance measures. The FDA recommends providing a description and flow chart, demonstrating how to randomize patients and readers into the different arms.
• Scoring Definition is an important component in the clinical assessment of a CADe device. The scoring process for the clinical studies should be consistent with the abnormalities marked by the CADe and the intended use of your device prior to initiating your evaluation. The scoring implementation should be integrated into the EDC and CRF including a procedure for determining the comparability between the reader’s interpretation and the ground truth (e.g., disease status)
• Integrative supportive reading centre, efficient EDC should store and process all study data in one central platform (Images, external data etc.).
  It is important to define hierarchic entry permissions and passwords to each role.

• Skilled Clinical, Data Management Team, planning of a reader study requires a good understanding of the obstacles that may arise throughout the study. Interactive visualizations and predictive analytics allow the clinical development team to rapidly explore clinical data for proactive and responsive decision making, in near real-time.

When multiple readers are involved in scoring, training in the use of the CADe device and rating interpretation scale is crucial and may support finer rating score.

 A skilled team will know how to combat challenges and avoid impediment in your clinical trial.

*The recommendations in this document are meant to guide you as you develop and lead the data collection steps of the clinical study; they are not meant to specify the full content or type of premarket submission that may be applicable to your device.