Missing data in trauma registries: A systematic review.

Area of Science:

• Medical Informatics
• Public Health
• Trauma Surgery

Background:

• Trauma registries are crucial for trauma systems, but their utility depends on data quality.
• Incomplete or missing data can significantly impact the reliability of research findings.
• Standardized reporting and management of missing data are essential for valid conclusions.

Purpose of the Study:

• To assess the reporting of data completeness in trauma registry studies.
• To evaluate the methods used to handle missing data in these publications.
• To identify gaps in current practices for managing missing data in trauma research.

Main Methods:

• A systematic review of trauma registry-based manuscripts published between January 2015 and March 2017.
• Searched MEDLINE, EMBASE, and CINAHL databases for relevant studies.
• Evaluated manuscripts based on recommendations for reporting and managing missing data, including quantification and analytical methods.

Main Results:

• Out of 539 included manuscripts, 38.6% did not mention data completeness, and 16.3% mentioned it without quantification.
• Only 4.8% quantified missingness for all variables; 88.5% lacked comparisons between complete and incomplete data cohorts.
• Among studies mentioning data management, 78.8% used complete case analysis, potentially introducing bias.

Conclusions:

• Most trauma registry studies exhibit poor reporting and minimal discussion of missing data.
• The prevalent use of complete case analysis, without addressing missingness, may bias research outcomes.
• Lack of standardized reporting and management of missing data compromises the validity of conclusions drawn from trauma registry research.