Data quality assurance and control in cognitive research: Lessons learned from the PREDICT-HD study

Area of Science:

• Neuroscience
• Clinical Research
• Data Science

Background:

• Huntington's disease (HD) research requires high-quality cognitive data for accurate prediction.
• Long-term observational studies like PREDICT-HD involve complex data collection from numerous participants.
• Ensuring data integrity is paramount for understanding disease progression and developing predictive biomarkers.

Purpose of the Study:

• To outline data quality control (QC) and quality assurance (QA) strategies for the cognitive assessments in the PREDICT-HD study.
• To detail examiner training, error correction methods, and data minimization techniques.
• To share key lessons learned to improve cognitive data accuracy in similar research.

Main Methods:

• Rigorous training and continuous evaluation of cognitive examiners.
• Systematic procedures for identifying and correcting data errors.
• Implementation of strategies to minimize potential data inaccuracies during collection.

Main Results:

• The PREDICT-HD cognitive core successfully implemented robust QC/QA protocols.
• Specific methods for examiner management and data error handling were effective.
• Five critical lessons were identified to enhance cognitive data reliability.

Conclusions:

• Effective data quality control is essential for the validity of neurobiological predictors in Huntington's disease research.
• Sharing best practices, such as those from PREDICT-HD, can significantly improve the quality of cognitive data in large-scale studies.
• Adherence to stringent QC/QA protocols minimizes errors and enhances the reliability of findings in prodromal Huntington disease research.