Inter-rater reliability of preventable death judgments. The Preventable Death Study Group.

Area of Science:

• Traumatology
• Medical Error Analysis
• Quality Improvement in Healthcare

Background:

• Assessing preventable deaths in trauma care is crucial for improving patient outcomes.
• Previous studies highlight variability in expert judgment regarding trauma preventability.
• Standardized methods are needed to enhance the reliability of preventability assessments.

Purpose of the Study:

• To evaluate the inter-rater reliability of preventable death judgments among expert panels in trauma cases.
• To identify factors influencing agreement and disagreement in preventability assessments.
• To inform future methodologies for reviewing trauma deaths.

Main Methods:

• 130 trauma deaths from Maryland (1986) were independently reviewed by multiple expert panels.
• Panels comprised trauma surgeons, neurosurgeons, and emergency physicians, with composition varying based on injury type (CNS vs. non-CNS).
• Reviewers classified deaths as not preventable, possibly preventable, probably preventable, or definitely preventable, with disagreements resolved through group discussion.

Main Results:

• Overall inter-rater reliability was low, with kappa values of 0.21 for non-CNS and 0.40 for CNS deaths.
• Agreement was highest when both autopsy and prehospital care reports were available.
• Disagreements primarily occurred between 'not preventable' and 'possibly preventable' categories, with higher agreement for early deaths and less severe injuries.

Conclusions:

• Current methods for judging trauma preventable deaths exhibit low reliability, necessitating methodological improvements.
• Availability of comprehensive patient data, including autopsy and prehospital reports, can enhance reliability.
• Further research is needed to develop robust and consistent approaches for assessing preventable trauma deaths.