Page 90 - JSOM Summer 2019

predefined search codes. This is a retrospective review of injuries sustained from explosives but higher proportion of
12
prospectively collected data within the registry. We searched gunshot wounds (GSWs), and were more likely to have sus-
our data set for all subjects with a documented prehospital tained injuries as part of Operation Enduring Freedom (Table
or ED intubation for inclusion into this analysis. We placed 1). The prehospital cohort had higher composite injury se-
subjects with documentation of both a prehospital and ED verity scores but lower abbreviated scores for the thorax re-
intubation into the prehospital category as they were likely gion (Table 2). Overall, a lower proportion of the prehospital
intubated (or status post attempted intubation) in the prehos- cohort survived to hospital discharge (76.4% versus 84.3%,
pital setting and redundant coding in the registry occurred. p < .001).
The US Army Institute of Surgical Research regulatory office
reviewed protocol H-16-005 and determined it was exempt On univariable logistic regression analysis, the prehospital
from institutional review board oversight. We obtained only cohort had lower odds of survival to hospital discharge (OR
deidentified data. 0.60, 95% CI 0.51–0.71). On multivariable analysis, con-
trolling for injury severity score, military operation, patient
category (US military, coalition, etc.), mechanism of injury,
Department of Defense Trauma Registry Description
sex, and age, the lower odds of survival to hospital discharge
The Department of Defense Trauma Registry (DODTR), for- persisted (OR 0.59, 95% CI 0.50–0.71).
merly known as the Joint Theater Trauma Registry (JTTR), is
the data repository for DoD trauma-related injuries. 16,17 The Head Injury Subgroup (AISBR1)
DODTR includes documentation regarding demographics, in-
jury-producing incidents, diagnoses, treatments, and outcomes There were 1486 patients with a head AIS ≥3, 449 of whom
of injuries sustained by US/non-US military and US/non-US were intubated prehospital and 1037 were intubated in the
civilian personnel in wartime and peacetime from the point of ED. Of patients intubated prehospital, 278 (61.9%) survived
injury to final disposition. The DODTR comprises all patients to hospital discharge versus 783 (75.5%, p < .001) of those
admitted to a Role 3 (fixed facility) or FST with an injury intubated in the ED. On univariable analysis, those intubated
diagnosis using the International Classification of Disease, prehospital had lower odds of survival to hospital discharge
Ninth Edition (ICD-9) between 800-959.9, near-drowning/ (OR 0.63, 95% CI 0.49–0.82). In multivariable analysis con-
drowning with associated injury (ICD-9 994.1) or inhalational trolling for age, sex, mechanism of injury, theater of operation
injury (ICD-9 987.9) and trauma occurring within 72 hours and AIS face, chest, abdomen, extremities, and external, the
from presentation. We defined the prehospital setting as any lower odds of survival persisted for patients intubated prehos-
location prior to reaching an FST or a CSH to include the Role pital (adjusted OR 0.49, 95% CI 0.49–0.82).
1 (point of injury, casualty collection point, battalion aid sta-
tion) and Role 2 (temporary limited-capability forward-posi- Discussion
tioned hospital inside combat zone without surgical support).
The registry categorization scheme considers a Role 2+ (or To the best of our knowledge, this is the first comparison of
variant with surgical support) to be the ED. adult trauma patients undergoing intubation in the prehospi-
tal versus ED setting in the combat environment. In this data
Analysis set, we found that those undergoing prehospital intubation
We performed all statistical analyses by using Microsoft Ex- had worse outcomes than those intubated in the ED, even
cel (version 10, Redmond, WA) and JMP Statistical Discovery after controlling for confounders. This suggests that patients
from SAS (version 13, Cary, NC). We compared study vari- in need of emergency airway intervention may derive benefit
ables between the subjects intubated in the prehospital setting from intubation delay until reaching the ED. The reasons for
versus the ED setting using a Student t-test for continuous vari- this are likely multifactorial.
ables expressed as means with standard deviations, Wilcoxon
rank sum test for ordinal variables expressed as medians and Unfortunately, the prehospital data do not have sufficient
interquartile ranges (IQRs), and χ test for nominal variables granularity to determine indications for intubation, which is
2
expressed as numbers and percentages. For binary outcomes, a limiting factor. In that regard, a frequent indication for in-
we used a logistic regression analysis to report ORs. tubation in the prehospital setting is failure of airway protec-
tion due to depressed mental state. Given this, we assessed
20
We performed a subgroup analysis of patients with significant the impact of intubation specifically among those patients
head injuries. To operationalize the Abbreviated Injury Scale with head injury as defined by an AIS ≥3 for the head body
9
(AIS) by body region (AISBR) as a binary variable, we dichot- region. We still found lower survival to hospital discharge
19
omized these data as either serious (≥3) or not serious (<3). 18,19 rates among those patients intubated in the prehospital setting
(61.9% versus 75.5% p < .001). Furthermore, the lower odds
of survival persisted in multivariable analysis controlling for
Results
potential confounders. This suggests that the association be-
Overall Analysis tween prehospital intubation and decreased survival applies
During the study period, there were a total of 38,769 encoun- in this brittle subgroup of patients with severe head injuries
ters in the DODTR. Our predefined search codes captured who are especially sensitive to hypoxia that may occur in the
28,222 (72.8%) of those subjects with our datas et. Within setting of airway intubation complications. Conversely, after
21
this data set, there was documentation of 4341 intubations: intubation a more rapid bag-tube rate may occur in which det-
1117 in the prehospital cohort and 3224 in the ED cohort. riment from hyperoxia can occur.

The prehospital cohort had a lower median age, comprised a The Role 1 and Role 2 (without FST attachment) settings
higher proportion of local forces, had a lower proportion of where the prehospital intubations took place are generally

88 | JSOM Volume 19, Edition 2 / Summer 2019