Comparison of Decision-Assist and Clinical Judgment of Experts for Prediction of Lifesaving Interventions.

Early recognition of hemorrhage during the initial resuscitation of injured patients is associated with improved survival in both civilian and military casualties. We tested a transfusion and lifesaving intervention (LSI) prediction algorithm in comparison with clinical judgment of expert trauma care providers. We collected 15 min of pulse oximeter photopletysmograph waveforms and extracted features to predict LSIs. We compared this with clinical judgment of LSIs by individual categories of prehospital providers, nurses, and physicians and a combined judgment of all three providers using the Area Under Receiver Operating Curve (AUROC). We obtained clinical judgment of need for LSI from 405 expert clinicians in135 trauma patients. The pulse oximeter algorithm predicted transfusion within 6 h (AUROC, 0.92; P < 0.003) more accurately than either physicians or prehospital providers and as accurately as nurses (AUROC, 0.76; P = 0.07). For prediction of surgical procedures, the algorithm was as accurate as the three categories of clinicians. For prediction of fluid bolus, the diagnostic algorithm (AUROC, 0.9) was significantly more accurate than prehospital providers (AUROC, 0.62; P = 0.02) and nurses (AUROC, 0.57; P = 0.04) and as accurate as physicians (AUROC, 0.71; P = 0.06). Prediction of intubation by the algorithm (AUROC, 0.92) was as accurate as each of the three categories of clinicians. The algorithm was more accurate (P < 0.03) for blood and fluid prediction than the combined clinical judgment of all three providers but no different from the clinicians in the prediction of surgery (P = 0.7) or intubation (P = 0.8). Automated analysis of 15 min of pulse oximeter waveforms predicts the need for LSIs during initial trauma resuscitation as accurately as judgment of expert trauma clinicians. For prediction of emergency transfusion and fluid bolus, pulse oximetry features were more accurate than these experts. Such automated decision support could assist resuscitation decisions, trauma team, and operating room and blood bank preparations.

How are you really feeling? A prospective evaluation of cognitive function following trauma.

BACKGROUND: Mild traumatic brain injury is associated with persistent cognitive difficulties. However, these symptoms may not be specific to the head injury itself. We sought to evaluate the prevalence of these symptoms in patients following trauma. METHODS: A prospective analysis of patients who were seen in the outpatient trauma clinic during a 20-month period and completed self-administered Rivermead Post-Concussion Symptoms Questionnaire was conducted. "Significant" difficulty with cognition was defined by two or more symptoms reported as severe or four or more symptoms reported as moderate. Head injury was defined as head Abbreviated Injury Scale (AIS) score greater than 0, including the diagnosis of concussion. Multivariable logistic regression was used to test associations between head injury, injury severity, sex, and age with significant cognitive difficulties, loss of work/school, and unmet physical, occupational, or psychological therapy needs. RESULTS: A total of 587 completed questionnaires were matched to trauma registry admissions (382 early, 111 mid, 86 late). The incidence of significant cognitive difficulties was 37% at less than 1 month, 40% at 1 month to 3 months, and 45% of patients at more than 3 months following injury. Head injury was not associated with increased odds for significant cognitive difficulties (adjusted odds ratio, 1.21; 95% confidence interval, 0.82-1.77; p = 0.3) There was no significant difference in symptoms in patients who carried a head injury diagnosis and those who did not. CONCLUSION: Cognitive problems occur frequently following injury even in the absence of a head injury diagnosis. Either mild traumatic brain injury is grossly underdiagnosed or these symptoms are not specific to postconcussive states and simply are the cognitive sequelae of traumatic injury. The reporting of moderate-to-severe symptoms suggests a need to better understand the effects of trauma on cognitive function and strongly suggests that services for these patients are badly needed to maximize cognitive function and return to preinjury quality of life. LEVEL OF EVIDENCE: Prognostic/epidemiologic study, level II.

Use of serum biomarkers to predict secondary insults following severe traumatic brain injury.

The management of severe traumatic brain injury (TBI) focuses on prevention and treatment of intracranial hypertension (ICH) and cerebral hypoperfusion (CH). Predicting which patients will develop these secondary insults is currently not possible. This study investigates the systemic manifestation of neuroinflammation and its role in helping to predict clinical deterioration following severe TBI. Patients with head Abbreviated Injury Severity greater than 3, age older than 14 years, "isolated" TBI, and placement of intracranial pressure monitor were prospectively enrolled. Serum was collected within 24 h and twice daily for 7 days. Measures of moderate and severe ICH (intracranial pressure >20 and >30 mmHg) and moderate and severe CH (cerebral perfusion pressure <60 and <50 mmHg) were compared with interleukin 8 (IL-8) and tumor necrosis factor α (TNF-α) levels drawn before periods of monitoring. An adjusted mixed-model analysis accounting for longitudinal correlations was applied. Sixty-eight patients were enrolled; 670 12-h periods of monitoring and 845 serum samples were available for analysis. Associations were found between serum levels of IL-8 and moderate and severe CH. Levels of TNF-α and severe ICH and CH were also correlated. Specificities of 81% to 95% were found for prediction of ICH and CH for TNF-α and CH for IL-8. Interleukin 8 and TNF-α demonstrate promise as candidate serum markers of impending ICH and CH. This suggests that we may be able to "predict" imminent events following TBI before clinical manifestations. Given the morbidity of ICH and CH, minimizing the effects of these secondary insults may have a significant impact on outcome and help guide decisions about timing of interventions.

Use of serum biomarkers to predict cerebral hypoxia after severe traumatic brain injury.

The management of severe traumatic brain injury (TBI) focuses on prevention and treatment of secondary insults such as cerebral hypoxia (CH). There are a number of biomarkers that are thought to play a part in secondary injury following severe TBI. This study evaluates the association between S100ß, neuron-specific enolase (NSE), and glial fibrillary acidic protein (GFAP), detected in the serum of severe TBI patients and CH as measured by brain tissue oxygen partial pressure (Pbo(2)). Patients with severe TBI were prospectively enrolled. Pressure times time (PTD; mm Hg*h), measuring the depth and duration of CH, was calculated for 12-h periods for episodes of moderate (Pbo(2) < 20 mm Hg) and severe (Pbo(2) < 15 mm Hg) CH, and compared to serum levels of S100ß, NSE, and GFAP drawn prior to periods of monitoring. An adjusted mixed model analysis was applied as was receiver operating characteristic (ROC) curve analysis. Of 76 patients enrolled, 24 had Pbo(2) monitoring. One hundred and thirty serum samples were matched with 12-h periods of monitoring. Significant associations were found in adjusted analyses between increasing serum levels of S100ß (coefficient=0.57, 0.56; p<0.001), NSE (coefficient=0.48, 0.52; p<0.001), and GFAP (coefficient=0.29, 0.30; p=0.003 and 0.002), and increasing PTD of moderate (Pbo(2)<20 mm Hg) and severe (Pbo(2)<15 mm Hg) CH. AUCs for the prediction of moderate and severe CH were 0.62 and 0.66 for S100ß, 0.55 and 0.71 for NSE, and 0.50 and 0.62 for GFAP, respectively. Specificities were between 76% and 90% for S100ß and NSE. S100ß, NSE, and GFAP demonstrate promise as candidate serum markers of impending CH. The fact that these biomarker elevations occur prior to the onset of clinical manifestations suggests that we may be able to predict imminent events following TBI. Given the morbidity of CH, early intervention and prevention may have a significant impact on outcomes and help guide decisions about the timing of interventions.

Relationship of serum and cerebrospinal fluid biomarkers with intracranial hypertension and cerebral hypoperfusion after severe traumatic brain injury.

BACKGROUND: There is little that can be done to treat or reverse the primary injury that occurs at the time of a traumatic brain injury (TBI). Initial management of the patient with severe TBI focuses on prevention of subsequent secondary insults, namely, intracranial hypertension (ICH) and cerebral hypoperfusion (CH). Currently, there is no reliable way to predict which patients will develop ICH and CH other than clinical acumen; therefore, indicators of impending secondary intracranial insults may be useful in predicting these events and allowing for prevention and early intervention. This study was undertaken to investigate the relationship of cytokine levels with intracranial pressure (ICP) and cerebral perfusion pressure (CPP) in patients with severe TBI. METHODS: Patients at the R Adams Cowley Shock Trauma Center were prospectively enrolled for a 6-month period. Inclusion criteria were older than 17 years, admission within the first 6 hours after injury, Glasgow Coma Scale<9 on admission, and placement of a clinically indicated ICP monitor. Serum and cerebrospinal fluid, when available, were collected on admission and twice daily for 7 days. Cytokine levels of interleukin (IL)-1ß, IL-6, IL-8, IL-10, and tumor necrosis factor (TNF)-α were analyzed by multiplex bead array assays. Hourly values for ICP and CPP were recorded, and means, minimum (for CPP) or maximum (for ICP) values, percentage time ICP>20 mm Hg (%ICP20) and CPP<60 mm Hg (%CPP60), and cumulative Pressure Times Time Dose (PTD; mm Hg·h) for ICP>20 mm Hg (PTD ICP20) and CPP<60 mm Hg (PTD CPP60) were compared with the serum and cerebrospinal fluid levels that were drawn before 12-hour time periods (PRE) and after 12-hour time periods (POST) of monitoring. RESULTS: Twenty-four patients were enrolled. In-hospital mortality was 12.5%, and good functional outcome was noted in 58%. Two hundred and seventy-five serum samples were taken and analyzed. IL-6 levels in the serum were found in the highest concentration of the cytokines measured. PTD ICP20 and PTD CPP60 were moderately correlated with increased PRE IL-8 levels (r=0.34, p<0.001; r=0.53, p<0.001). PTD ICP20 was also correlated with PRE TNF-α levels (r=0.27, p<0.001) as was PTD CPP60 (r=0.25, p<0.001). POST IL-8 levels were found to be correlated with PTD ICP20 (r=0.46, p<0.001) and PTD CPP60 (r=0.54, p<0.001). POST TNF-α was associated with PTD ICP20 (r=0.45, p<0.001). PTD CPP60 was also moderately correlated with POST TNF-α levels (r=0.26, p<0.001). When comparing patients with good versus poor outcome, median daily serum IL-8 levels were associated with poor outcome. CONCLUSIONS: IL-8 and, to a lesser extent, TNF-α demonstrated the most promise in this study to be candidate serum markers of impending ICH and CH. The clinical relevance of this is the suggestion that we may be able to predict impending secondary insults after TBI before the clinical manifestation of these events. Given the known morbidity of ICH and CH, early intervention and prevention may have a significant impact on outcome. This becomes even more important when decisions must be made about timing of interventions. Increased levels of IL-8 and TNF-α in the serum during episodes of ICH and CH imply there are significant systemic effects of these events. These serum biomarkers are promising as diagnostic targets. In addition, further study of the precise role of these molecules may have significant implications for inflammatory system manipulation in the management of severe TBI.

Association of CSF biomarkers and secondary insults following severe traumatic brain injury.

BACKGROUND: Management of severe traumatic brain injury (TBI) focuses on mitigating secondary insults. There are a number of biomarkers that are thought to play a part in secondary injury following severe TBI. Two of these, S100ß and neuron-specific enolase (NSE), have been extensively studied in the setting of neurological injury. This pilot study was undertaken to investigate the relationship of S100ß and NSE to clinical markers of severity and poor outcome: intracranial hypertension (ICH), and cerebral hypoperfusion (CH). METHODS: Patients at the R Adams Cowley Shock Trauma Center were prospectively enrolled over an 18-month period. Inclusion criteria were: age > 18, admission within the first 6 h after injury, Glasgow Coma Scale (GCS) < 9 on admission, isolated TBI, and placement of an intraventricular catheter (IVC). Patients were managed according to an institutional protocol based on the Brain Trauma Foundation Guidelines. CSF was collected from the IVC on admission and twice daily for 7 days. S100ß and NSE levels were analyzed by ELISA. CSF levels drawn before (PRE) and after (POST) 12-h time periods were compared to percentage time intracranial pressure (ICP) > 20 mmHg (% ICP(20)) and cerebral perfusion pressure (CPP) < 60 mmHg (% CPP(60)), and cumulative "Pressure times Time Dose" (PTD) for episodes of ICP > 20 mmHg (PTD ICP(20)) and CPP < 60 mmHg (PTD CPP(60)). Statistical analysis was performed using the Student's t test to compare means and non-parametric Wilcoxon statistic to compare ranked data. Linear regression methods were applied to compare levels of S100ß and NSE with ICP and CPP(.) RESULTS: Twenty-three patients were enrolled. The cohort of patients was severely injured and neurologically compromised on admission (admission GCS = 5.6 ± 3.1, Injury Severity Score (ISS) = 31.9 ± 10.6, head Abbreviated Injury Scale (AIS) = 4.4 ± 0.7, Marshall score = 2.6 ± 0.9). Elevated levels of S100ß and NSE were found in all 223 CSF samples analyzed. ICH was found to be associated with PRE and POST S100ß levels when measured as % ICP(20) (r = 0.20 and r = 0.23, P < 0.01) and PTD ICP(20) (r = 0.35 and r = 0.26, P < 0.001). POST increasing NSE levels were weakly correlated with increasing PTD ICP(20) (r = 0.17, P = 0.01). PRE S100ß levels were associated with episodes of CH as measured by % CPP(60) (r = 0.20, P = 0.002) and both PRE and POST S100ß levels were associated with PTD CPP(60) (r = 0.24 and r = 0.23, P < 0.001). PRE and POST NSE levels were also associated with episodes of CH as measured by % CPP(60) (r = 0.22 and r = 0.18, P < 0.01) and PTD CPP(60) (r = 0.20 and r = 0.21, P < 0.01). CONCLUSIONS: In this preliminary analysis, S100ß levels were associated with ICH and CH over a full week of ICP monitoring. We also found associations between CH and NSE levels in CSF of patients with severe TBI. Our results suggest that there is an association between levels of ICH and CH and these biomarkers when measured before episodes of clinically significant secondary insults. These markers of neuronal cell death demonstrate promise as both indicators of impending clinical deterioration and targets of future therapeutic interventions.