An investigation of discrepancies in outcome reporting and selective reporting bias in interrupted time series studies of health interventions: a methodological study
Methodological study investigates selective reporting bias in interrupted time series studies of health interventions, highlighting significant discrepancies in outcome reporting.
Key Points
This study aims to investigate discrepancies between reported outcomes and initial protocols in interrupted time series studies.
Systematically searched 22 databases for published protocols of interrupted time series studies from 2010 to 2022.
Identified outcomes from the protocol and results reports, categorizing discrepancies in reporting.
Classified results based on completeness, significance, and direction of effects.
24% of outcomes in the protocol were not reported in the results report, affecting 60% of protocols.
52% of assessed outcomes had discrepancies, impacting 74% of studied protocols.
Only 28% of discrepancies were justified in the results report.