Multivariable Prognostic Analysis in Traumatic Brain Injury: Results from The IMPACT Study

Public Health Sciences, University of Edinburgh Medical School, Edinburgh, United Kingdom.
Journal of Neurotrauma (Impact Factor: 3.71). 03/2007; 24(2):329-37. DOI: 10.1089/neu.2006.0035
Source: PubMed


We studied the prognostic value of a wide range of conventional and novel prognostic factors on admission after traumatic brain injury (TBI) using both univariate and multivariable analysis. The outcome measure was Glasgow Outcome Scale at 6 months after injury. Individual patient data were available on a cohort of 8686 patients drawn from eight randomized controlled trials and three observational studies. The most powerful independent prognostic variables were age, Glasgow Coma Scale (GCS) motor score, pupil response, and computerized tomography (CT) characteristics, including the Marshall CT classification and traumatic subarachnoid hemorrhage. Prothrombin time was also identified as a powerful independent prognostic factor, but it was only available for a limited number of patients coming from three of the relevant studies. Other important prognostic factors included hypotension, hypoxia, the eye and verbal components of the GCS, glucose, platelets, and hemoglobin. These results on prognostic factors will underpin future work on the IMPACT project, which is focused on the development of novel approaches to the design and analysis of clinical trials in TBI. In addition, the results provide pointers to future research, including further analysis of the prognostic value of prothrombin time, and the evaluation of the clinical impact of intervening aggressively to correct abnormalities in hemoglobin, glucose, and coagulation.

Download full-text


Available from: Ewout W Steyerberg,
1 Follower
182 Reads
  • Source
    • "of secondary insults or structural abnormalities on imaging, and laboratory abnormalities to be predictive of eventual outcome [9]. However, clinical course after TBI is often unpredictable , and most prognostic models are unable to incorporate information gleaned from the patients' treatment and recovery. "
    [Show abstract] [Hide abstract]
    ABSTRACT: In patients with severe traumatic brain injury, increased intracranial pressure (ICP) is associated with poor functional outcome or death. Hypertonic saline (HTS) is a hyperosmolar therapy commonly used to treat increased ICP; this study aimed to measure initial patient response to HTS and look for association with patient outcome. Patients >17 years old, admitted and requiring ICP monitoring between 2008 and 2010 at a large urban tertiary care facility were retrospectively enrolled. The first dose of hypertonic saline administered after admission for ICP >19mmHg was recorded and correlated with vital signs recorded at the bedside. The absolute and relative change in ICP at 1 and 2h after HTS administration was calculated. Patients were stratified by mortality and long-term (≥6 months) functional neurological outcome. We identified 46 patients who received at least 1 dose of HTS for ICP>19, of whom 80% were male, mean age 34.4, with a median post-resuscitation GCS score of 6. All patients showed a significant decrease in ICP 1h after HTS administration. Two hours post-administration, survivors showed a further decrease in ICP (43% reduction from baseline), while ICP began to rebound in non-survivors (17% reduction from baseline). When patients were stratified for long-term neurological outcome, results were similar, with a significant difference in groups by 2h after HTS administration. In patients treated with HTS for intracranial hypertension, those who survived or had good neurological outcome, when compared to those who died or had poor outcomes, showed a significantly larger sustained decrease in ICP 2h after administration. This suggests that even early in a patient's treatment, treatment responsiveness is associated with mortality or poor functional outcome. While this work is preliminary, it suggests that early failure to obtain a sustainable response to hyperosmolar therapy may warrant greater treatment intensity or therapy escalation.
    Injury 09/2014; 45(12). DOI:10.1016/j.injury.2014.08.041 · 2.14 Impact Factor
  • Source
    • "Many factors have been reported to influence the prognosis of TBI, including age, gender, severity of injury, co-morbidities, concomitant use of anticoagulants, secondary insults, the initial Glasgow Coma Scale (GCS) score, the motor score, pupil reactivity, the type of lesion visualized on brain computed tomography (CT) scan, changes in intracranial pressure (ICP) and blood levels of specific proteins. Two prognostic scores for TBI patients on admission have been developed on the basis of the large, prospective IMPACT and CRASH databases [4,5]. However, these scores are complex, require numerous data and are not easy to perform in the emergency unit. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Introduction The aim of this study was to evaluate the prognostic value of optic nerve sheath diameter (ONSD) measured on the initial brain computed tomography (CT) scan for intensive care unit (ICU) mortality in severe traumatic brain injury (TBI) patients. Methods A prospective observational study of all severe TBI patients admitted to a neurosurgical ICU (over a 10-month period). Demographic and clinical data and brain CT scan results were recorded. ONSD for each eye was measured on the initial CT scan. The group of ICU survivors was compared to non-survivors. Glasgow Outcome Scale (GOS) was evaluated six months after ICU discharge. Results Seventy-seven patients were included (age: 43 ± 18; 81% males; mean Injury Severity Score: 35 ± 15; ICU mortality: 28.5% (n = 22)). Mean ONSD on the initial brain CT scan was 7.8 ± 0.1 mm in non-survivors vs. 6.8 ± 0.1 mm in survivors (P < 0.001). The operative value of ONSD was a good predictor of mortality (area under the curve: 0.805). An ONSD cutoff ≥ 7.3 had a sensitivity of 86.4% and a specificity of 74.6% and was independently associated with mortality in this population (adjusted odds ratio 95% confidence interval: 22.7 (3.2 to 159.6), P = 0.002). There was a relationship between initial ONSD values and six-month GOS (P = 0.03). Conclusions ONSD measured on the initial brain CT scan is independently associated with ICU mortality rate (when ≥ 7.3 mm) in severe TBI patients.
    Critical care (London, England) 03/2013; 17(2):R61. DOI:10.1186/cc12589 · 4.48 Impact Factor
  • Source
    • "Several reports concerning the cognitive prognosis after TBI have emerged in the literature from retrospective convenience samples using univariate analysis and without a clear control of missing (drop out) cases (Christensen et al., 2008; Novack, Alderson, Bush, Meythaler, & Canupp, 2000). Some prospective studies investigated the cognitive recovery over time after the TBI (Bayen et al., 2012; Dikmen, Machamer, Powell, & Temkin, 2003; Sigurdardottir, Andelic, Roe, & Schanke, 2009), but the independent association among the variables previously reported to be associated with mortality or morbidity evaluated by Glasgow Outcome Scale (Martins et al., 2009; Murray et al., 2007; Perel, Edwards, Wentz, & Roberts, 2006) and the cognitive prognosis remains to be investigated. Our hypothesis was that variables classically associated with TBI prognosis (like Marshall CT classification, Glasgow Coma Scale [GCS], pupils examination, and admission blood glucose levels) could also be used to predict the cognitive outcome of severe TBI victims. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Objectives: Traumatic brain injury (TBI) is a main cause of mortality and morbidity. Association studies between hospitalization variables and cognitive impairment after TBI are frequently retrospective, including non-consecutive patients showing variable degrees of TBI severity, and poor management of missing (drop out) cases. Methods: We assessed prospectively the demographic and hospitalization variables of 234 consecutive patients with severe TBI (admission Glasgow Coma Scale [GCS] ≤8) and determined their independent association with cognitive performance in a representative sample (n = 46) of surviving patients (n = 172) evaluated 3 (±1.8) years after hospitalization. Results: In all, 85% of patients were male and the mean age was 34 (SD ±13) years. The education level was 9 (±4.7) years. As expected, education and age showed a moderately to strong linear relationship with the cognitive performance in 14 of 15 neuropsychological tests (R coefficient = 0.6-0.8). The cognitive test scores were not independently associated with gender, admission GCS, associated trauma, and Marshal CT classification. Admission-elevated blood glucose levels and the presence of sub-arachnoid haemorrhage were independently associated with lower scores on Rey Auditory Verbal Learning retention and Logical Memory-I tests, respectively. Conclusions: After correction for education and age distribution, the variables that are commonly associated with mortality or Glasgow Outcome Scale including admission pupils' examination, Marshal CT Classification, GCS, and serum glucose showed a limited predictive power for long-term cognitive prognosis. Identification of clinical, radiological, and laboratory variables as well as new biomarkers independently associated with cognitive outcome remains an important challenge for further work involving severe TBI patients.
    Journal of Neuropsychology 11/2012; 8(1). DOI:10.1111/jnp.12000 · 2.49 Impact Factor
Show more