Identification of preventable trauma deaths: confounded inquiries?

Area of Science:

• Medical quality assurance
• Trauma care research
• Patient safety evaluation

Background:

• Published methods for identifying preventable trauma deaths often overlook confounding factors.
• Reliability of consensus systems in trauma death reviews requires further investigation.

Purpose of the Study:

• To compare the reliability of three different consensus systems for identifying preventable trauma deaths.
• To assess the impact of discussion and reassignment on consensus agreement in case reviews.

Main Methods:

• Three five-member panels reviewed 20 non-central nervous system trauma fatalities using distinct consensus methods (independent judgment, discussion-first, and judgment-discussion-reassignment).
• Evaluated reliability using Kappa concordance index and consensus agreement rates.
• Assessed the effect of discussion on equivocal case designation.

Main Results:

• All methods demonstrated low Kappa concordance (0.20–0.40).
• Only one of 11 deaths deemed preventable by any panel was agreed upon by all three panels.
• Consensus agreement varied significantly across methods (20%–45%), with panel B (discussion-first) showing higher agreement than panel C (judgment-discussion-reassignment) (p<0.03).
• Discussion reduced equivocal case designation in panel C from 30% to 5%.

Conclusions:

• Individual case review methods for identifying preventable trauma deaths are unreliable and yield inconsistent results.
• Subjective case review should not be used to measure institutional quality of patient care.
• Objective evaluation methods focusing on patient population outcomes are recommended for assessing institutional performance.