Page 50 - NNS 2014 Program & Abstract Book

antibiotic administration for the treatment of pneumonia was shorter
in the early group (p = 0.04). Klebsiella species were the most com-
mon pathogens in both groups.
Early tracheostomy decreased the antibiotic period for the treatment
of pneumonia in patients with severe TBI who underwent decom-
pressive craniectomy. Early tracheostomy did not reduce total time of
mechanical ventilation, ICU stay, pneumonia incidence, and GOS.
Key words
intensive care unit, tracheostomy, traumatic brain injury
A1-04
INFLAMMATION IN TRAUMATIC BRAIN INJURY WITH
HEMORRHAGIC SHOCK
Lei, Z.G., McKinley, T., Rodgers, R.B., Territo, P.R., Meyer, J.,
Peters, J., Xu, Z.C.
Indiana University School of Medicine, Indianapolis, USA
Patients with traumatic brain injury (TBI) or hemorrhage shock (HS)
frequently develop the systemic inflammation response syndrome
(SIRS). SIRS is a primary risk factor associated with developing or-
gan dysfunction and poor prognosis. TBI and HS are common injury
components in multiply injured patients and often occur together. It is
not clear how combined TBI and HS affect the pathogenesis of sys-
temic inflammation. In this study, we developed a rat model of
combined TBI/HS and quantified serum changes in damage associated
molecular patterns that occurred due to TBI with or without HS.
Subsequently, we quantified changes in liver AMPK as a surrogate for
inflammation-mediated organ dysfunction.
TBI was induced in adult Sprague Dawley rats using controlled
cortical impact (CCI) method. HS was induced by controlled hem-
orrhage via femoral artery to a mean arterial pressure of 40 mmHg
(maintained 60 min). Mitochondrial DNA (mtDNA) fragments were
measured by PCR and high-mobility group protein 1 (HMGB1) was
measured using ELISA from 3 hours to 72 hours after injury. AMPK
was measured from the liver after sacrifice using ELISA.
Severe ipsilateral cortical injury was observed 24 h after TBI, and the
presence of HS did not affect the severity of injury. Serum levels of
mtDNA gradually increased up to 72 hours after injury in TBI speci-
mens. In TBI/HS specimens, significant increases in mtDNA occurred
3 hrs post injury and returned to baseline at 24 hours before a second
significant increase at 72 hours post injury. Serum HMGB1 levels change
in a pattern similar to those of mtDNA. Liver parenchymal AMPK levels
were not affected in TBI specs and decreased 60% in TBI/HS specimens.
The present study established an animal model of TBI + /- HS in
rats. The above results suggest that HS potentiates inflammation
caused by TBI. In addition, remote organ energetic changes occurred
in animals sustaining combined TBI/HS. More studies are needed to
elucidate the inflammation pathways and develop potential agents to
block this pathological process.
Key words
AMPK, hemorrhagic shock, HMGB1, mtDNA, TBI
A1-05
DEVELOPMENT OF A PORTABLE CEREBRAL MICRO-
DIALYSIS PLATFORM FOR AUTOMATED INLINE MULTI-
ANALYTE DETECTION SYSTEM
Das, C., Agarwal, A., Nguyen, C., Wang, G., Ledden, B., Kumar, S.
SFC Fluidics, Fayetteville, USA
An estimated 1.7 million traumatic brain injury (TBI) cases are reported
every year in the United States with 15% of these being severe enough
to require invasive monitoring. There is an urgent need for development
of a bed-side, microdialysis system with in-built analyzer which will
help in rapid and continuous monitoring of the state of health of a
patient with severe TBI. The current state-of-the-art involves sample
collection, labeling, storage and batch analysis that can take hours, if
not days, which precludes results from being effectively used for
clinical decision making. This protocol not only requires trained prac-
titioners, but also is prone to errors in handling tiny amounts of dial-
ysate. Moreover, current commercial instruments are limited to the
analysis of small molecule metabolic biomarkers; these instruments do
not detect large molecule biomarkers such as proteins. MD Analyzer ,
being developed by SFC Fluidics, is an advancement over current
microdialysis-based diagnostics for severe TBI, because all steps of the
analysis will occur at bed side; clinically actionable results will be
provided every 15 minutes for metabolic biomarkers, lactate (0–8
mmol/L), glutamate (1–400 lmol/L), pyruvate (0–250 lmol/L), and
every hour for protein biomarker, S100B (20–200 ng/mL), detection.
The first generation prototype described here integrates different in-
novative technologies like a non-mechanical electrochemical pumping
system ePump , Quickconnect fluidics coupling, electrochemical
sensors for small molecule detection and a flow through Enzyme
Linked Immunosorbent Assay (ELISA) for protein detection. This fully
automated system moves the perfusate through a standard microdialysis
probe, collects the dialysate and analyzes it in near-real time with little
or no user intervention. Data presented will be detection of different
concentration of biomarkers in spiked artificial cerebrospinal fluid.
Availability of this nurse-friendly fully integrated system that auto-
matically collects analyzes and reports the dynamic changes in con-
centrations of clinically relevant biomarkers will significantly
ameliorate the patient care in Neurointensive Care Units (NICU).
Key words
ELISA, instrumentation, microdialysis, real-time detection, S100B
A1-06
VALIDATION OF THE IMPACT PROGNOSTIC MODELS
Madden, L.K.1,2, Blozis, S.A.3, Keachie, K.1, Shahlaie, K.1,
Berman, R.F.1, DeVon, H.A.4
1UC Davis, Department of Neurosurgery, Sacramento, US
2UC Davis, Betty Irene Moore School of Nursing, Sacramento, US
3UC Davis, Department of Psychology, Davis, US
4University of Illinois at Chicago, College of Nursing, Chicago, US
Several prognostic models have been developed for traumatic brain in-
jury (TBI) but have limited applicability due to development from small
datasets. The International Mission for Prognosis and Analysis of Clin-
ical Trials in TBI (IMPACT) models, derived from several clinical trial
datasets, were developed for prediction of mortality versus survival or
unfavorable versus favorable outcome at 6-months post-injury. In order
to confirm utility across centers, external validation is necessary. The
purpose of this study was to evaluate the performance of three IMPACT
models for prediction of 6-month post-injury mortality and dichotomized
neurologic outcome (unfavorable/favorable) using data collected from a
Level I trauma center registry. We analyzed admission data on 354
consecutive adult ( ‡ 16 years) patients that sustained blunt TBI with
Glasgow Coma Scale sum score £ 12 (moderate or severe TBI). IM-
PACT model predictions of mortality (Extended Glasgow Outcome
Scale [GOS-E] = 1) versus survival (GOS-E = 2-8) and unfavorable
(GOS-E = 1–4) versus favorable outcome (GOS-E = 5–8) were compared
to the actual observed classifications using logistic regression. The ability
A-18