Validation of safety outcomes in routinely collected data:Lessons learned from a multinational postapproval safety study
Key Points - Trial-based case definitions assume close patient monitoring that captures both clinical and subclinical disease manifestations. Adapting such definitions as reference standards to validate cases in observational studies of routinely collected data is therefore challenging. - In validation studies, trial-based case definitions should be calibrated to reflect routine medical care. - When multiple databases are involved in a validation study, database- and country-specific differences in the availability and completeness of the required information should be taken into account. - Re... Mehr ...
Key Points - Trial-based case definitions assume close patient monitoring that captures both clinical and subclinical disease manifestations. Adapting such definitions as reference standards to validate cases in observational studies of routinely collected data is therefore challenging. - In validation studies, trial-based case definitions should be calibrated to reflect routine medical care. - When multiple databases are involved in a validation study, database- and country-specific differences in the availability and completeness of the required information should be taken into account. - Researchers should also consider more than one validation approach to ensure cases are accurately captured; additionally, procedures for quantifying uncertainty should be explored. - Results from validations of case definitions in routinely collected data should be appropriately interpreted in light of how the case definitions were derived.