Asymmetry of Bispectral Index (BIS) in severe brain-injured patients treated by barbiturates with unilateral or diffuse brain injury.

OBJECTIVE: Bispectral index (BIS) may be used in traumatic brain-injured patients (TBI) with intractable intracranial hypertension to adjust barbiturate infusion but it is obtained through a unilateral frontal electrode. The objective of this study was to evaluate differences in BIS between hemispheres in two groups: unilateral frontal (UFI) and diffuse (DI) injured. PATIENTS AND METHODS: Prospective monocenter observational study in 24 TBI treated with barbiturates: 13 UFI and 11 DI. Simultaneous BIS and EEG was recorded for 1h. Goal of monitoring was a left BIS between 5 and 15. Biases in BIS were considered as clinically relevant if greater than 5. Differences in biases were interpreted from both statistical (Mann-Whitney test) and clinical points of view. RESULTS: Mean BIS in the two hemispheres remained in the same monitoring range. There were statistic and clinical differences in some values in the two groups of patients (15% of bias greater than I5I in UFI group and 10% in DI group). BIS monitoring allowed the adequate number of bursts/minutes to be predicted in 18 patients and did not detect an overdosage in 2. CONCLUSIONS: While asymmetric BIS values in TBI patients occur whatever the kind of injury, they were not found to be clinically relevant in most of these heavily sedated patients. Asymmetrical BIS monitoring might be sufficient to monitor barbiturate infusion in TBI provided that the concordance between BIS and EEG is regularly checked.

[Changes in intracranial pressure in severe head injured patients in hemodialysis]. / Variations de la pression intracrânienne chez le traumatisé crânien grave en hémodialyse.

Two cases of acute post-traumatic renal failure in severely head injured patients are reported. An increase in intracranial pressure (ICP) was shown up by continuous monitoring during haemodialysis: it was more important during conventional haemodialysis than during continuous arteriovenous haemofiltration. Although this effect is well known experimentally, few cases of continuous ICP pressure monitoring have been reported in head injury patients undergoing haemodialysis. The increase in ICP is explained in the dog as a result of blood-brain differences in urea concentration and osmolality leading to an increase in cerebral spinal fluid volume and cerebral tissue swelling. If dialysis is necessary in these patients, it should be carried out early and progressively, the patient's ICP being monitored continuously.

[Measuring severity in trauma resuscitation]. / Mesure de la gravité en réanimation traumatologique.

ISS is the most widely used anatomic trauma severity index. This scale uses clinical judgement. It is therefore a subjective method, the validity of which is somewhat questionable. The aim of this study was to compare the clinical index ISS with a new non-specific score; the "simplified acute physiology score" (SAPS). Five-hundred trauma patients were checked retrospectively. Were excluded small traumas which did not require ICU hospitalization. ISS were all calculated by the same specialist and SAPS was drawn out of the initial blood sample examination. The sensitivity (proportion of true positive) and the specificity (proportion of true negative) were calculated for the two scores at all different cut-off points. "Receiver operating characteristic curves" (ROC) were drawn and their areas were compared by means of the Hanley test. The best Youden index, i.e. the fewest false positives for the most true positives, was also calculated. Mean ISS was 23.3 +/- 9.2. Youden index was 0.1; sensitivity and specificity were respectively 57 and 52% for the cut-off points. Mean SAPS was 8.7 +/- 4.3. Youden index was 0.3 and sensitivity and specificity were respectively 68 and 62%. ROC curve area for SAPS was 0.69 +/- 0.02. ROC curve area for ISS was 0.56 +/- 0.03 (p = 0.0001). In trauma, SAPS seemed to be a better predictor of outcome than ISS.