Abstract P180: The Risk of Infectious and Septic Complications from Trauma are Age-Associated: An Analysis of the National Trauma Database
INTRODUCTION: Understanding the immune response to injury may allow us to better characterize tolerance to injury and shock.
OBJECTIVE: To evaluate the relationship between age and infectious morbidity associated with injuries by investigating if the risk of pneumonia, bacteremia and acute respiratory distress syndrome (ARDS) differ across age groups in patients with traumatic injuries.
METHODS: Patients were identified by using the National Trauma Database (2000–2004) which is the largest aggregation of trauma registry data. This database represents over 75% of the level I and level II trauma centers in the nation. Patients were included if they were at least 5 years old and did not include subjects with burns nor unidentified mechanisms of injury. Multivariate logistic regression adjusted for gender, injury severity score, days of ventilation, chest abbreviated injury severity, and mechanism were done to compare age groups. Children between 5 to 12 years old were used as the comparison group.
RESULTS: There were 863,393 patient visits analyzed for this study. The mortality rate was 4.6%. The patient characteristics are summarized as averages for age 39.9 years, 66% male, and 60% Caucasian. The complication rates of pneumonia 1.6%, bacteremia 0.1%, and ARDS 0.5% were noted. Table 1⇓ lists the adjusted odds ratios for pneumonia. The adjusted odds ratio for bacteremia and ARDS have similar values and each age group was observed to have a statistically significant increased risk.
CONCLUSION: There is an ordinal risk of pneumonia, bacteremia, and ARDS with age among injured patients after controlling for multiple confounding variables. These findings suggest that the immune response to trauma differs between age groups and further suggests that there is a neuroendocrine axis of response.
LIMITATIONS: Post-traumatic complications are likely under-reported and therefore the observed associations are likely diminished, representing conservative estimations.