Identifying types and causes of errors in mortality data in a clinical registry using multiple information systems

Area of Science:

• Healthcare Quality Improvement
• Clinical Data Management
• Critical Care Medicine

Background:

• In-hospital mortality registration is crucial for assessing healthcare quality.
• Existing clinical quality registries may contain registration errors, affecting data reliability.
• The National Intensive Care Evaluation (NICE) registry is a key source for intensive care unit (ICU) quality metrics.

Purpose of the Study:

• To identify and categorize errors in in-hospital mortality registration within the NICE registry.
• To determine the frequency and causes of these registration errors.
• To provide recommendations for improving the accuracy of mortality data in clinical registries.

Main Methods:

• Comparative analysis of mortality data between the NICE registry and a national insurance claims database.
• On-site investigations at eleven Intensive Care Units (ICUs) to examine error types and causes.
• Systematic identification and classification of registration errors.

Main Results:

• A total of 255 errors were identified in the NICE registry's in-hospital mortality data.
• Software malfunctions accounted for approximately 80% of the identified errors.
• The remaining 20% of errors comprised manual transcription mistakes and failures in recording outcome data.

Conclusions:

• In-hospital mortality data in clinical registries can be prone to errors, impacting its utility as a quality indicator.
• Software issues are a significant contributor to registration inaccuracies.
• Recommendations include rigorous software verification and addressing manual data entry challenges to enhance data integrity.