RWD-Cockpit

RWD-Cockpit evaluates the quality of real-world data by translating dataset characteristics into quantifiable metrics to provide preliminary quality assessment for observational research and regulatory decision-making.

Key Features:

• Quantifiable metrics: Translates characterization of specific quality parameters into numerical metrics for dataset evaluation.
• Systematic scoring: Systematically scores datasets based on defined quality variables to produce comparative quality assessments.
• Seven quality variables: Assesses manageability (access and publication status), complexity (univariate, multivariate, and longitudinal data), sample size, privacy and liability (privacy rules), accessibility (data access granularity), periodicity (update frequency), and standardization (adherence to technical or metadata standards).
• Descriptor mapping: Associates descriptors with each variable to define dataset-specific characteristics used in scoring.
• Customizable variables: Allows selection of variables to tailor the scoring framework while retaining an internal standard for comparability.
• Output quality identifiers: Produces output scores that provide an initial assessment of RWD quality.
• Cross-domain applicability: Applies the scoring framework to molecular, phenotypical, and social datasets.
• Validation on sleep datasets: Validated using both newly generated and publicly available sleep-related real-world datasets.

Scientific Applications:

• Preliminary RWD assessment for observational studies: Provides initial quality identifiers to support population-based observational research and regulatory evidence generation.
• Data quality improvement: Identifies variables and dataset characteristics that can guide efforts to enhance RWD quality.
• Standardized comparative evaluation: Enables application of a common preliminary evaluation standard across molecular, phenotypical, and social data sets.
• Validation and benchmarking: Supports validation of RWD sources and benchmarking using sleep-related datasets (new and public).

Methodology:

Characterization of quality parameters is translated into quantifiable metrics and datasets are systematically scored across seven defined variables with associated descriptors to produce output quality scores.