Comparative Effectiveness of Inhospital Trauma Resuscitation at a French Trauma Center and Matched Patients Treated in the United States

Abstract

Objective: 

The objective of this paper is to compare mortality outcomes between patients treated at a trauma center in France and matched patients in the United States.

Background: 

Although trauma systems in France and the United States differ significantly in prehospital and inhospital management, previous comparisons have been challenged by the lack of comparable data.

Methods: 

Coarsened exact matching identified matching patients between a single center trauma database from Lyon, France, and the National Trauma Data Bank (NTDB) of the United States. Moderate to severely injured [injury severity score (ISS) > 8] adult patients (age ≥ 16) presenting alive to level 1 trauma centers from 2002 to 2005 with blunt or penetrating injuries were included. After matching patients, multivariate regression analyses were performed to determine difference in mortality between patients in Lyon and the NTDB.

Results: 

A total of 1043 significantly injured patients were presented to the Lyon center. Matching eligible patients with complete records were sought from 219,985 patients in the NTDB. The unadjusted odds of mortality at the Lyon center was 2.5 times higher than that of the NTDB [95% confidence interval (CI) = 2.18–2.98]. However, the Lyon center received patients with higher ISS, lower Glasgow Coma Score (GCS), and lower systolic blood pressure (SBP) (all P < 0.001). After 1:1 matching, 858 patient pairs were produced, and the odds of mortality became equivalent [odds ratio (OR) = 1.3, 95% CI = 0.91–1.73]. Similar results were found in multiple subset analyses.

Conclusions: 

Trauma patients admitted to a single French trauma center had an equal chance of survival compared with similarly injured patients treated at US trauma centers.

In Brief

The objective of this paper is to compare mortality outcomes between severely injured patients treated at a trauma center in Lyon, France, and patients with equivalent injuries in the United States. A total of 858 French patients were matched with comparable patients present in the US National Trauma Data Bank and after multivariate risk adjustment, there was no difference in survival (OR = 1.3, 95% CI = 0.91–1.73) between the 2 groups.