- Email: [email protected]
Cerebral oxygenation and systemic trauma related factors determining neurological outcome after brain injury
Journal of Clinical Neuroscience (2000) 7(3), 226–233 © 2000 Harcourt Publishers Ltd DOI: 10.1054/ jocn.1999.0202, available online at http://www.idealibrary.com on
Clinical study
Cerebral oxygenation and systemic trauma related factors determining neurological outcome after brain injury Javier Fandino1 MD, Reto Stocker2 MD, Sonja Prokop2 RN, Otmar Trentz 2 MD, Hans-Georg Imhof1 MD 1
Department of Neurosurgery, Division of Trauma, University Hospital Zurich, Switzerland and 2Department of Surgery, Division of Trauma, University Hospital Zurich, Switzerland
Summary We examined the relationship between clinical and radiological findings, cerebral oxygenation patterns during intensive care management, presence of systemic trauma related injuries and severity of illness in 50 patients (age: 32.3 ± 12 years, GCS: 8 ± 4) who were rescued from the accident scene within a 30 min period after trauma. Presence of systemic injuries was quantified using the Injury Severity Score (ISS) and severity of illness was scored using the Acute Physiology and Chronic Health Evaluation (APACHE II). Cerebral oxygenation parameters included continuous monitoring of jugular bulb oxygen saturation (SjvO2) for 12 840 h, and 2323 periodical blood sampling for measurement of arteriovenous differences in oxygen content (AVDO2), arteriovenous difference of lactate (AVDL) and lactate oxygen index (LOI). Fifteen patients (30%) presented with anisocoria or non-reacting pupils. Diffuse lesions on computed tomography (CT) were found in 34% of the patients and in 66% a mass lesion was removed. The mean ISS was 28 ± 15.3 and 34 patients (68%) had an APACHE II score between 20 and 29 (mean 24 ± 15). No statistically significant association between age (P = 0.45), gender (P = 0.83), initial Glasgow Coma Score (GCS) (P = 0.43), episodes of cerebral perfusion pressure (CPP) < 70 mm Hg (P = 0.8), ISS (P = 0.28), pupillary abnormalities (P = 0.57), initial CT findings (P = 0.74), APACHE II scores (P = 0.36) and outcome could be demonstrated. The number of SjvO2 desaturations (< 60%) was the only statistically significant factor associated with outcome (P = 0.05). The percentage of patients with poor neurological outcomes (GOS 1–3) was 38% in patients with no or one desaturation episode, and 57.6% in those with multiple desaturations. In conclusion, in patients who are resuscitated early and quickly transferred to the hospital, the number of SjvO2 desaturations during intensive care management might be associated with outcome more strongly than other clinical and radiological features. © 2000 Harcourt Publishers Ltd Keywords: severe head injury, Injury Severity Score, APACHE II, intensive care, jugular bulb oxygen saturation, outcome
INTRODUCTION
Table 1 Demographic and trauma-related factors
Several studies have demonstrated that physiological factors like systemic hypotension, intracranial hypertension, arterial hypoxia and hypocapnia are associated with poor outcome after brain injury.1–5 Introducing new techniques into clinical practice, such as jugular bulb oxygen saturation (SjvO2), allowed a better understanding of cerebral oxygenation patterns during intensive care management of head injured patients.6–9,10,11 The incidence of SjvO2 desaturation episodes and, more recently, low brain tissue oxygen pressure values (PbtO2 below 15 torr), were reported to be associated with a poor neurological outcome.12,13 Brain injury must be considered and managed as a systemic condition in which many extracranial factors may influence outcome. Besides clinical and radiological findings, well established scores of systemic injuries, such as the Injury Severity Score (ISS), and scores denoting the severity of systemic illness, like the Acute Physiology and Chronic Health Evaluation score (APACHE II), have been routinely used for assessing the prognosis of trauma patients who receive intensive care management.14–16 The purpose of the present study was to examine the relationship between clinical and radiological findings, systemic trauma related factors such as ISS and APACHE II, cerebral oxygenation parameters, and outcome in severe head injured patients who were resuscitated within a 30 min period after injury and rapidly transferred to the hospital.
Total of patients Age (mean ± SD) Sex (ratio M:F) Male (%) Female (%) Glasgow Coma Scale score (mean ± SD) 3–8 (%) 9–12 (%) 13–15 (%) Trauma mechanism Fall Car Motorcycle Assault/gunshot Sport Bicycle Others ISS (mean ± SD) Type of injury Open/isolated (%) Open/multitraumatised (%) Closed/isolated (%) Closed/multitraumatised (%) APACHE II (mean ± SD) Duration of ICU management (mean ± SD) Glasgow Outcome Scale score (mean ± SD)* D (%) VS and SD (%) MD and GR (%)
Received 23 February 1999 Accepted 25 May 1999 Correspondence to: Javier Fandino MD, Department of Neurosurgery, University Hospital Zurich, Frauenklinkstrasse 24, Zurich, Switzerland. Tel: +41–1–2552660; Fax: +41–1–2554505; E-mail: [email protected]
226
50 32.3 ± 12 years 2.3 35 (70) 15 (30) 8±4 26 (52) 14 (28) 10 (20) 16 (38%) 10 (22%) 7 (14%) 5 (10%) 2 (4%) 2 (4%) 7 (14%) 28 ± 15.3 24 (48) 1 (2) 5 (10) 20 (40) 24 ± 5 28.4 ± 15.2 days 4±1 9 (18) 16 (32) 25 (50)
Abbreviations: SD: standard deviation; ISS: Injury Severity Score;3,4 APACHE: Acute Physiology and Chronic Health Evaluation;17 ICU: Intensive Care Unit; *: Outcome 3 months after trauma (D: death, VS: vegetative state; SD: severe disability; MD: moderate disability; GR: good recovery).
Cerebral oxygenation and outcome after brain injury 227
CLINICAL MATERIAL AND METHODS Patient population, clinical assessment and management Fifty consecutive patients with severe brain injuries admitted to the Trauma Intensive Care Unit (TICU) during a 32 month period were included in this study. The demographic characteristics are shown in Table 1. As entry criteria patients had to be rescued and assessed in the scene of the accident within the first 30 min after injury, and immediately transferred to our hospital by helicopter or ambulance equipped with mechanical ventilator and intensive care equipment. The mean age was 32.3 ± 12 years (range 17–60). Seventy percent of the patients were men (n: 35) and 30% were women (n: 15). The mechanism and cause of trauma was verified using the resuscitation and referral reports. Eighteen patients (40%) suffered head injury in transport related accidents (car, motorcycle and bicycle). None of the bicycle riders (4%) was using a helmet. A fall was the trauma mechanism in 16 patients (32%). Violence related episodes were the cause of head injuries in 5 patients (10%). Sports related injuries were found in 2 patients (4%) and other causes in 2 patients (4%). Patients with fatal gunshot injuries or brain death suspicion on admission were not included in the study. Twenty-five patients (50%) had open head injuries, most of whom (48%) had isolated brain injuries. On the other hand, 25 patients (50%) suffered closed brain injuries, the majority of which were multitraumatised (40%). A total of 29 patients (58%) presented with isolated brain injuries and 21 patients (42%) were multitraumatised. Glasgow Coma Score17 (GCS) was assessed at the scene of the accident and on admission to the hospital. Considering the aim of this study and the possible multiple factors influencing the outcome, we did not define severity of brain injury in terms of GCS. Correlation of GCS and outcome was analysed as a score and not as a category of severity of brain injury. The majority of the patients (n: 38, 76%) were intubated at the scene of the accident after neurological evaluation and came to the emergency room paralysed and sedated. Pupillary size and response were documented at the scene of the accident, emergency room and periodically after admission in the TICU. Oxygen saturation (SaO2) was immediately monitored in the rescue scene and, when possible, blood gas analyses were made in order to rule out a hypoxic episode prior to resuscitation or admission into the hospital. All patients were classified according to the ISS.14,15 After admission in the TICU, patients were assigned specific diagnostic categories and scored using the APACHE II which provides a general measure of severity of disease.16 APACHE II uses a point score based upon initial values of 12 routine physiologic measurements, age and previous health status. At TICU admission, any emergency operation was recorded for later mortality risk prediction. Our management protocol has been described elsewhere.18,19 Therapeutic tools included circulatory monitoring with a pulmonary artery catheter (Swan-Ganz), invasive mean arterial pressure monitoring (MAP), cerebrospinal fluid (CSF) drainage, hyperventilation under SjvO2 monitoring, osmotherapy with mannitol and mild hypothermia. Sixteen patients (32%) underwent barbiturate coma the thiopentone dose was adjusted according to bedside; electroencephalography (EEG) registration (6 bursts/min). Corticosteroids were not included in our treatment protocol and no antiepileptic drugs were administrated phrophylactically. Continuous intracranial pressure (ICP) monitoring was performed on all patients. Induced hypertension with cathecolamines and slight hypervolaemia was maintained in order to keep cerebral perfusion pressure (CPP) over 70 mm Hg. The most common ICP device used was a subdural catheter (49%). Ventriculostomy was performed in 32% of the patients. In 2% of the patients ICP was © 2000 Harcourt Publishers Ltd
monitored with infrared devices (Camino Laboratories,® San Diego CA). A wake up procedure was indicated if no ICP increase above 15 mm Hg within 24 h happened under normothermia and normoventilation, if the amount of CSF drainage was below 80 ml/24 h, if SjvO2 and AVDL values were normal, and if no signs of intracranial hypertension was evident on the computed tomography (CT) scan. ICP monitoring was discontinued as soon as neurological assessment was possible and no therapeutic interventions were necessary for 24 h. Radiological evaluation All patients underwent radiological evaluation with skull projections and CT immediately after admission to the hospital. Initial CT scans were evaluated according to the Traumatic Coma Data Bank (TCDB):20 categories 1–4 (diffuse injuries); 5 (evacuated mass lesion); and 6 (non-evacuated lesion). During ICU management, patients underwent CT follow up within 48 h after admission, before the wake up procedure if ICP increased or became intractable with critical CPP, or if unexplained deterioration of cerebral oxygenation parameters appeared. Cerebral angiography was performed in patients with presence of atypical posttraumatic subarachnoid haemorrhage (tSAH) and suspicion of ruptured aneurysm, traumatic arteriovenous fistulae (cervical, dural, cortical or carotid-cavernous) or venous sinus injuries. Bedside SjvO2 monitoring and measurements of AVDL To monitor SjvO2, a No. 5.5 French fiberoptic oxygen double lumen catheter was inserted percutaneously into the internal jugular vein through a No. 6 French introducer sheath according to the technique describe in detail elsewhere.21,22 The position of the catheter was controlled radiologically with anteroposterior and lateral skull projections. The catheter for measurement SjvO2 was placed on the right side unless a unique unilateral lesion in the left hemisphere was present, the left jugular circulation was demonstrated to be dominant, or if thrombosis or thrombi in the right jugular vein were documented in the colour coded duplex sonography. On-line printouts of SjvO2 curves were available. Every effort was made to start the monitoring as early after admission as possible. Pre-insertion and in vivo calibration was performed after insertion and every 6 h thereafter. If SjvO2 changed abruptly or if technical measurement problems were suspected, in vivo calibration was immediately repeated and documented. SaO2 and haemoglobin concentration was, if possible, kept constant throughout the ICU management, depending on presence of lung or chest injuries. Normal SjvO2 values were considered between 60% and 75%. We defined a desaturation episode as SjvO2 lower than 60%. Causes of artifactual measurement were ruled out before considering cerebral hypoperfusion using an algorithm described elsewhere.10 SjvO2 greater than 75% suggested luxury perfusion.19 Two patient groups were considered for the analysis of outcome: patients with no or one desaturation episode, and patients with multiple desaturation episodes. If desaturation was suspected based on the fiberoptic measurements, blood gas analysis was done to confirm the values. AVDO2 was calculated from the difference between arterial and jugular vein oxygen content. Arteriovenous difference of lactate (AVDL) was calculated subtracting jugular vein from arterial (systemic) lactate concentration values. AVDL and SjvO2 (and AVDO2) were measured in order to identify simultaneous episodes of luxury perfusion and ischaemia. AVDL values of 0.2 µmol/L or greater were interpreted as increased cerebral lactate production due to an ischaemia/infarction episode. Lactate oxygen index LOI (LOI = –AVDL/AVDO2) was calculated according to Robertson et al.10 A LOI of 0.08 or greater was interpreted as an Journal of Clinical Neuroscience (2000) 7(3), 226–233
228 Fandino et al. Table 2 Demographic characteristics and outcome
Mean age (SD) M:F ratio (% men) Mean initial GCS (SD) Mean Injury Severity Score (SD) Mean APACHE II (SD) Initial CT findings (TCDB classification) Patients with diffuse injuries: categories 1–4 (%) Patients with mass lesions: categories 5–6 (%) Patients with pupillary abnormalities (%) Patients with CPP < 70 mm Hg (%) No episode One episode Two or more episodes Number of SjvO2 desaturation episodes/Percentage of patients who presented SjvO2 desaturation episodes Number of patients (%)
GOS 4–5
GOS 2–3
GOS 1
P-value*
30.5 (12.8) 7.3 (88) 10 (3.4) 25.6 (5.8) 21.1 (3.8)
34.7 (12.2) 1.6 (62) 7 (3.8) 29.7 (7.2) 25.2 (2.2)
33.2 (13.4) 0.5 (33) 7 (3.6) 31.4 (4.4) 24.8 (5.6)
0.45 0.83 0.46 0.28 0.36 0.74
9 (18) 16 (34) 8 (16)
5 (10) 11 (20) 4 (8)
3 (6) 6 (12) 3 (6)
11 (22) 6 (12) 8 (16) 93/48
10 (20) 1 (2) 5 (10) 134/56
5 (10) 2 (4) 2 (4) 11/71
25 (50)
16 (32)
9 (18)
0.57 0.8
0.05
Abbreviations: GOS: Glasgow Outcome Scale score 3 months after trauma;15 SD: standard deviation; M: male; F: female; GCS: Glasgow Coma Scale score; APACHE: Acute Physiology and Chronic Health Evaluation;17 CT: computed tomography; TCDB: National Institutes of Health Traumatic Coma Data Bank;20 SjvO2: jugular bulb oxygen saturation; *Significance considering GOS groups.
ischaemia/infarction pattern. All parameters (ICP, CPP, mean arterial pressure (MAP), SjvO2, AVDO2, AVDL, SaO2, SvO2, PaO2 and PCO2) were documented in a ‘cerebral metabolism’ database at least every 6 h. Lactate measurements were done in the laboratory (TDx FLx lactate analyzer, Abbott Laboratories, Abbott Park, IL, USA). Blood gas analyses were performed in the unit (ABL System, Radiometer, Radiometer Medical A/S, Copenhagen, Denmark).
patients (44%) had GR, 10 patients (20%) were MD and 6 patients (12%) were SD. There was no statistically significant difference in age (P = 0.45) or gender (P = 0.83) distribution in relation to GOS (Table 2). Patients who had GOS 4 or 5 (GD or MD) were in 84% of the cases males and were 30.5 years (SD 12.8). Patients with a GOS of 2 or 3 (MD or SD) were in 62% of the cases males (m:f ratio = 1.6) with an age of 34.7 ± 12.2 years. Finally, 33% of patients who died (GOS 1) were males (m:f ratio = 0.5) with an age of 33.2 ± 13.4 years.
Determination of outcome The Glasgow Outcome Score (GOS) was used to quantify the severity of the neurological deficits.23 Outcome was determined 3 months and one year after injury by personnel who were unaware of the cerebral oxygenation data obtained during the intensive care management. The patients were routinely controlled at the neurological department in a special out-patient unit for brain injured patients. A standard battery of memory tests, social integration assessment and neurological examination were included in the evaluation. Statistical analysis Summary data are reported as the mean ± standard deviation. Categorical variables of assessment of brain injury and jugular vein saturation values were analysed using Chi-square whereas differences in mean were tested using a one-way analysis of variance. The significant level used to determined statistical significance was 0.05. RESULTS Outcome after three months and one year after trauma Outcome evaluation 3 months after brain injury showed that 8 patients (16%) had good recovery (GR), 18 patients (36%) were moderately disabled (MD), 14 patients (28%) were severely disabled (SD), 1 patient (2%) was in a vegetative state (VS) and 9 patients (18%) died (D) (Tables 1 and 2). Six patients died because of intractable intracranial hypertension and three because of medical complications associated with deterioration of CPP. One patient, who was in VS, died eight months after discharge from the hospital because of pulmonary complications. GOS, after one year follow up, showed an improved outcome in 20 patients (40%): 22
Journal of Clinical Neuroscience (2000) 7(3), 226–233
Initial clinical and neurological assessment The average of initial GCS was 8 ± 4. Twenty-six patients (52%) presented with initial GCS ≤ 8, 14 patients (28%) had GCS between 9 and 12, and 10 patients (20%) had GCS between 13 and 15. As shown in Table 2, patients with an initial GOS of 4 or 5 (GR or MD) had GCS of 10.3 ± 3.4. Patients with lower GOS (SD, VS or D) had a similar initial GCS. Patients with GOS 2 or 3 (SD or VS) had an initial GCS of 7.6 ± 3.8, and patients who died (GOS 1) had a similar GCS (7.4 ± 3.6). In our series, initial GCS was not significantly significant associated with outcome (P = 0.46). Initial pupillary examination revealed that 70% (n: 35) of the patients had normal and reacting pupils. Thirty percent of the patients (n: 15) presented with anisocoria or non-reacting pupils. We found no statistically significant relationship between pupillary abnormalities and outcome after 3 months (P = 0.34). At least in our series, there were more patients who presented with anisocoria and had GOS 4 or 5 (n: 8) than patients who had GOS 2 or 3 (n: 4) or died (n: 3). Initial computed tomography and outcome Diffuse lesions (TCDB Categories 1–4) were found in 17 patients (34%) and mass lesions (TCDB Categories 5–6) were present in 33 patients (66%). Nineteen patients (38%) underwent craniotomy and removal of a haematoma immediately after admission. The number of patients who presented with mass lesions was twice as many as patients who had diffuse injury of the brain in all GOS groups (Table 2). Sixteen patients (32%) with mass lesions, compared to 9 patients (18%), had a GOS 4 or 5 after 3 months. Eleven patients (22%) with initial mass lesions on CT, compared to 5 patients (10%), were severely disabled or persistently vegetative in the outcome evaluation. In the group of
© 2000 Harcourt Publishers Ltd
Cerebral oxygenation and outcome after brain injury 229
Fig. 1 Injury Severity Score (ISS)2,3 and GOS 3 months after trauma. No statistically significant correlation could be demonstrated (P = 0.28). GOS: 1: death; 2: vegetative state; 3: severe disabled; 4: moderate disabled; 5: good recovery.
Fig. 3 Acute Physiology and Chronic Health Evaluation (APACHE II17) and GOS three months after trauma. No statistically significant correlation could be demonstrated (P = 0.36). GOS: 1: death; 2: vegetative state; 3: severe disabled; 4: moderate disabled; 5: good recovery.
lower than 10. There were 3 patients in each APACHE II category group. The mean APACHE II in patients with GOS 4 and 5 was 21 ± 3.8, in patients with GOS 2 and in patients who died the mean was 24.8 ± 5.6. APACHE II was not a mortality predictor in this series and, as shown in Figure 3, the relationship between APACHE II and GOS was not statistical significant (P = 0.36). CPP deterioration and neurological outcome
Fig. 2 Relation of Acute Physiology and Chronic Health Evaluation (APACHE II17) and mortality after brain injury.
patients who died, 6 had mass lesions (12%) and 3 (6%) presented with diffuse injuries. In this series, initial CT findings did not correlate significantly with outcome (P = 0.74). Injury severity score and outcome The mean initial ISS was 28 ± 15. The majority of the patients had an ISS between 20–39. Twenty-eight patients (56%) presented with an ISS between 20 and 29, and 10 patients (20%) had an ISS between 30 and 39. Only 5 patients (10%) had an ISS greater than 40. Thirty-eight percent of the patients included in this study underwent extracranial emergency operation under ICP monitoring to ensure adequate CPP intraoperatively. Figure 1 shows the relation between ISS and the outcome 3 months after injury. Patients with GOS 4 or 5 had an ISS of 25 ± 6, patients with GOS 2 or 3 had a ISS of 29 ± 7, and patients who died had an ISS of 31 ± 4. In the group of ISS < 20 (n: 7), the majority of the patients had a GOS 2 or 3 (30%). The majority of patients who died (GOS 1) had an ISS between 30 and 39. The association between ISS and GOS was not statistically significant in this series (P = 0.28). Acute physiology and chronic health evaluation (APACHE II) We classified our patients into three APACHE score category groups (Fig. 2). Thirty-four patients (68%) had a score between 20 and 29, 11 patients (22%) had a score between 10 and 19, and 5 patients had an APACHE of more than 30. No patient had a score © 2000 Harcourt Publishers Ltd
Table 2 summarises the number of patients presenting episodes of CPP less than 70 mm Hg. In a total of 26 patients (52%) CPP could be maintained over 70 mm Hg, 9 patients (18%) one episode of CPP deterioration which required therapeutic intervention could be documented. Finally, 15 patients (30%) had two or more persistent episodes of CPP > 70 mm Hg which required any therapeutic intervention. No statistically significant association could be demonstrated between CPP deterioration and neurological outcome. SjvO2 desaturation episodes and neurological outcome A total of 12 840 h of on-line SjvO2 monitoring during mechanical ventilation were collected. In addition 2323 measurements of AVDO2 and AVDL at 6 h intervals were performed. The mean duration of intensive care management in this series was 28.4 ± 15.2 days. The long duration of ICU care was due mainly to the fact that 32% of the patients underwent barbiturate coma. The duration of continuous SjvO2 monitoring ranged between 2 and 35 days (mean, 12 ± 3 days). All patients underwent jugular vein cannulation for SjvO2 monitoring and AVDL measurements before starting hyperventilation. In our series, hyperventilation without signs of anaerobic metabolism due to ischaemia/infarction signs (AVDL ≥ 0.2 µmol/L and/or SjvO2 < 60%) was possible in only 13 of 50 patients (26%). In 37 patients (74%) hyperventilation was contraindicated due the risk of ischaemia. A total of 238 SjvO2 desaturation episodes were detected or confirmed in the periodical jugular bulb blood sampling (10.2% from the total of periodical measurements). Desaturation episodes were most common within the first 48 h after injury (68%). The incidence of desaturation and percentage of patients who presented one or more desaturation episodes was higher in patients with poor outcomes (Table 2). A higher percentage of desaturations was observed among the patients who died (71%), compared to those patients who had GOS 2 or 3 (56%) or patients with GOS 4-or 5 (48%). As presented in Figure 4, the percentage of patients with poor neurological outcomes (GOS 1–3) was 38% in patients with no or one desaturation episodes, and 57.6% in those Journal of Clinical Neuroscience (2000) 7(3), 226–233
230 Fandino et al.
patients (66%). Seventeen patients did not show abnormal LOI during the intensive management. The majority of the patients had five or less episodes (measurements) of abnormal LOI (27 patients). Five patients presented more than five episodes of LOI of 0.08 or greater. No statistically significant association with outcome was found in patients with more than 5 episodes of LOI ≥ 0.08 (GOS 1–3: 1 patient, 4–5; 6 patients) or those with five or less episodes of abnormal LOI (GOS 1–3: 12 patients, 4–5: 14 patients). Complications of jugular vein catheterisation
Fig. 4 Correlation between SjvO2 desaturation episodes and GOS three months after trauma (P = 0.05). GOS: 1: death; 2: vegetative state; 3: severe disabled; 4: moderate disabled; 5: good recovery.
with multiple (2 or more) desaturation episodes. The percentage of patients with favourable outcomes (GOS 4–5) with no or one desaturation episode was 60%, compared to those with multiple desaturations (41.8%). Mortality in the group of patients who experienced multiple desaturations was higher (6 of 9 patients who died during intensive care management had multiple desaturations) than in patients who had no or one desaturation. The incidence of jugular venous desaturation episodes (none or one compared to multiple) was the only factor associated with neurological outcome in this study (P = 0.05). Simultaneous SjvO2 and AVDL abnormalities
The incidence of carotid puncture could not be clearly quantified retrospectively in our series; nevertheless, the introduction of Doppler and colour coded duplex guided punctures diminished the incidence of this complication throughout the study period. Subcutaneous haematoma after carotid puncture did not cause any major complications such as airway obstruction, diminished flow in the carotid artery or acute ICP elevations. The tips of the 105 catheters used in the 50 patients were cultured and no bacteriological growth could be demonstrated in any specimen. As soon as the puncture site showed inflammatory signs (started to redden, or serious fluid was detected) the catheter and introducer were replaced. Thrombosis of the internal jugular vein was documented only in a young female patient (Fig. 5). This woman died on day 12 after brain injury suffering intractable generalised brain oedema with uncontrollable electrolyte disorders associated with a severe coagulopathy. An immune complex disorder was diagnosed (positive Raji cell assay) and was considered to be a clinical factor associated with thrombus formation. DISCUSSION
In 47 patients (94%) abnormal lactate values (AVDL ≥ 0.2 µmol/L) were measured. A total of 366 AVDL measurements (15.7%) were found to be abnormally high as a consequence of brain lactic acidosis. Only three patients had normal AVDL throughout the intensive care management. Table 3 shows SjvO2 values in relation to GOS in patients with AVDL ≥ 0.2 µmol/L. The simultaneous measurement of SjvO2 in the 47 patients who presented one or more episodes of AVDL ≥ 0.2 µmol/L was 70% ± 9. Thirty-three patients (66%) presented luxury perfusion signs by means of abnormal AVDL and SjvO2 ≥ 75%; nevertheless, no significant association with poor outcome could be demonstrated (11 patients had GOS 1–3 and 22 patients had GOS 4–5). All patients who presented one or more episode of brain lactic acidosis during the course of the intensive care management also had one or multiple SjvO2 desaturations. No statistically significant differences were found in the analyses of the associations between the different SvjO2 subgroups and outcome in patients with abnormal AVDL. Lactate oxygen index A lactate oxygen index (LOI = –AVDL/AVDO2) of 0.08 or greater was found in 110 of 2326 measurements (0.5%) in 33 out of the 50
Initial clinical and radiological findings Predictive factors after brain injury have been extensively studied in the last two decades.3–5,24 In the early 1980s several studies led to the consensus that factors predicting poor outcome after brain injury include the presence of an intracranial haematoma, increasing age, abnormal motor responses, impaired or absent eye movements or pupil light reflexes, early hypotension, hypoxaemia or hypercarbia, and elevation of ICP over 20 mm Hg despite artificial ventilation.4 In addition, other studies reported poor outcome in patients with ICP instability and hypotension,20 prolonged hyperventilation5 and CPP under 70 mm Hg.1 Based on the experience gained in the last two decades, specialists in intensive care, paramedics and other disciplines other than neurosurgery have assumed a crucial role preventing secondary brain injuries. Diffusion of concepts and protocols in the medical and paramedical community, and the introduction of new monitoring techniques, have been important accomplishments in our country over the last decade.18,19 Switzerland, with an area of 41.293 sq. km, has traditionally been a pioneer in terrestrial and air patient transportation. Since most of our patients are rescued and transported immediately after trauma, we decided to analyse demographic, clinical
Table 3 Outcome in patients who presented one or more episode of brain lactic acidosis (AVDL ≥ 0.2 µmol/L) distributed according to SjvO2 values (P = NS). SjvO2 Group 61%–74% ≥ 75% ≤ 60%
No. measurements*
No. patients
Mean SjvO2
GOS 1–3
GOS 4–5
366 (15.7%) 110 (4.7%) 122 (5.2%)
47 (94%) 33 (66%) 47 (94%)
70 ± 9% 79 ± 6% 56 ± 6%
19 (40.5%) 11 (34.4%) 19 (40.5%)
28 (59.5%) 22 (66.6%) 28 (59.5%)
Abbreviations: GOS:15 Glasgow Outcome Score (percentage of number of patients in each SjvO2 group); *in parenthesis: percentage of total number of measurements from the total of 2326 periodical measurements done in 50 patients.
Journal of Clinical Neuroscience (2000) 7(3), 226–233
© 2000 Harcourt Publishers Ltd
Cerebral oxygenation and outcome after brain injury 231
Fig. 5 Longitudinal sonographic view of an obstructing thrombus (arrow head) in the internal jugular vein (ijv). The flow cannot be visualised beyond the entrance of the catheter in the vein (arrows).
and radiological factors influencing outcome in 50 consecutive patients resuscitated within a 30 min period after injury. Miller et al. published a study describing the factors important in predicting outcome based on 225 consecutive severely head injured patients.4 The authors reported that approximately half of the patients were referred from other hospitals and 20% were seen by a neurosurgeon within the first hour after injury. The understanding of pathophysiological mechanisms in diffuse axonal injury and cerebral ischaemia helped introduce the concept of the concept of secondary brain damage.25–27 Since then, efforts have made in this decade to emphasise the importance of early resuscitation, surgical treatment and intensive care management. Undoubtedly, outcome predictors have changed, especially in patients who are transferred to the hospital in an ultra-short period of time. In our series, age did not significantly influence outcome; nevertheless, we should mention that our patients’ ages were homogeneous and did not exceed 60 years (32 ± 12 years). Severity of initial radiological findings and initial GCS did not significantly differ in the three outcome groups (P = 0.46); nevertheless, one must consider that most of our patients are already intubated and mechanically ventilated at the scene of the accident, which does not allow for additional reliable clinical assessments. Patients who do not obey commands, or who have a GCS of 8 or less, were usually sedated and paralysed before admission. This policy might influence the predictive value of GCS; nevertheless, it enables better haemodynamic and respiratory control. Another clinical factor, which has been described as an outcome predictor in severely head injured patients is the pupil light response.4,28 According to Ropper, it is a horizontal shifting that gives rise to ipsilateral pupillary abnormalities by overstretching the peripheral pupillary pathways within the oculomotor nerve and does not result, as was previously thought, from the compression caused by downward shifting.29 Pupillary abnormalities were found to have prognostic value in a study with 55 comatose individuals with supratentorial mass lesions, including 15 patients © 2000 Harcourt Publishers Ltd
with brain injury.28 In the Richmond series, however, correlations between bilateral absence of the pupil response to light, impaired or absent oculocephalic responses and a poor outcome were present in both surgical (P < 0.001) and non-surgical patients (P < 0.001). In our series, 8 out of the 15 patients who presented with anisocoria survived having a GOS 4 or 5. Only 3 patients who presented with anisocoria died. In this study, we found no statistically significant relationship between pupillary abnormalities and outcome after 3 months (P = 0.34). Based on a prospective study of over 750 initial CT scans in patients with severe head injury, The National Institutes of Health Traumatic Coma Data reported that the status of the mesencephalic cisterns and the degree of midline shift are strongly associated with high ICP and death.30 In our study, in spite of the fact that no statistically significant differences were found when comparing the different outcome groups in relation to the initial CT classification, the presence of a mass lesion seemed to be an ominous predictor of poor outcome. Thirty-three patients with mass lesions died, were vegetative or were severely disabled compared to 8 patients in whom diffuse injuries were initially diagnosed. Finally, we have to highlight that our results are based exclusively on the evaluation of CT findings in terms of TCDB classification. The fact that other factors such as midline shift, compression of basal cisterns and presence of subarachnoid blood were not independently evaluated might constitute a limitation of this study. Severity of systemic illness: ISS and APACHE Considering that brain injury must be approached as a systemic condition, we analysed the ISS and APACHE along with the initial clinical and radiological findings in order to quantify the degree of systemic lesions and the patient’s systemic condition. ISS is a method for numerically describing the overall severity of injury;15 it can be applied to persons who have sustained injury to more than one area of the body as well as to those with isolated injuries. The Journal of Clinical Neuroscience (2000) 7(3), 226–233
232 Fandino et al.
ISS correlates substantially better with mortality than does the Abbreviated Injury Scale (AIS) rating for the single most severe injury.14 In spite of the fact that ISS was introduced in the early 1970s, it has not been used routinely in the evaluation of brain injured patients. Since multitrauma is frequent in patients with brain injury (42% of the patients in our series were multitraumatised and 58% presented with isolated brain injuries), we believe that this score is useful not only in the assessment of systemic injuries, but also as a factor which might influence outcome. Miller et al. reported that despite the association between multiple injury and some systemic insults, there was no significant relationship between the presence of multiple injuries and outcome.4 More recently, Gospinath et al. classified a series of patients using the ISS, but found no statistically significant relationship between ISS and outcome after brain injury.12 The ISS ‘lethal dose or score’ for 50% of the patients is 40 for ages 15–44, 29 for ages 45–64 and 20 for ages 65 or older.15 The mean ISS in our study was 28 ± 15.3. In patients who died, the ISS was 31.4 ± 4.4 and in patients who survived with a GOS 4 or 5, the ISS was lower (25.6 ± 5.8). No statistically significant difference, nevertheless, could be observed. According to the Baltimore and Birmingham series, a mortality rate of approximately 20% would be expected in our series, which might coincide with the mortality rate in our study (18%). In addition, ISS has proven to be exponentially related to the grade of disability after trauma. Patients with an ISS of approximately 18 were reported to be severely disabled after trauma; in our study, patients severely disabled (GOS 2 or 3) presented with a much higher ISS (29.7 ± 7.2). These observations suggest that ISS might predict mortality but not disability in brain injured patients. The basis for the development of APACHE was the hypothesis that the severity of acute disease can be measured by quantifying the degree of abnormality from multiple physiologic variables.16 The number of physiological measurements in the original APACHE introduced in 1981 have been reduced from 34 to 12. The new version of APACHE or APACHE II include the following parameters: temperature, MAP, heart rate, respiratory rate, oxygenation measurements, arterial pH, serum sodium, serum potassium, serum creatinine, haematocrit, white blood count, GCS, age and a so called chronic health point for emergency and elective postoperative patients. In a multicentric validation study, Knaus et al. evaluated APACHE II in 5815 ICU admissions and demonstrated a direct relationship between high scores and the hospital death rates.16 For each five point increase in APACHE II, there was a significant increase in death rate; nevertheless, the overall risk of hospital death varied according to the disease. In addition, non-operative and postoperative patients have different death rates in the different APACHE II groups. In our study, the mean APACHE II score was 24 ± 5. According to APACHE II, a mortality rate between 30% and 40% would be expected in our series (in both operative and non-operative patients); nevertheless, a lower hospital death rate of 18% was, in fact, observed. On the other hand, APACHE II scores did not differ significantly in the three outcome groups (GOS 1: 24.8 ± 5.6, GOS 2–3: 25.2 ± 2.2 and GOS 4–5: 21.1 ± 3.8). One speculative explanation of the discordance between the expected and observed mortality rates could be the diversity of diagnoses included in the APACHE II study. Among the patients included in the study, only 2.3% had suffered head injury (both non-operative and operative) and only 4.8% were admitted because of multiple trauma. Our results might suggest that APACHE II is not a reliable predictor of hospital death in head injured patients, even if they presented as multitraumatised. Regardless of this fact, as the authors of the APACHE study themselves admit, prognostic estimates are still only estimates and devotion of the ICU personnel in the care of patients will always play the most important role in outcome.
Journal of Clinical Neuroscience (2000) 7(3), 226–233
Cerebral oxygenation patterns and outcome Since its introduction in the early 1980s, SjvO2 monitoring has become an important tool for understanding oxygenation patterns after brain injury.9 Complications resulting from this technique have been reported to be few and minor.31 SjvO2 enables the diagnosis of jugular vein desaturation episodes, brain lactic acidosis and reliable estimates of cerebral blood flow (CBF).10,11,32 The occurrence of jugular venous desaturation has been shown to be associated with poor neurological outcome.12 In our study we demonstrated that the percentage of patients who presented desaturation episodes was higher in patients who died (71%) than in patients who had GOS 2–3 (56%) or GOS 4–5 (48%). In addition, the percentage of patients with favourable outcomes who had no or one desaturation episode was significantly higher than among those who had multiple episodes. These observations, which highlight the importance of assessing the global cerebral oxygen supply and demand as an outcome predictor, were recently reinforced by studying the relationship between focal brain tissue oxygen pressure (PbtO2) and outcome after severe head injury.13 We believe that understanding cerebral oxygenation patterns and quickly detecting hypoperfusion states have led to better outcome rates in severely head injured patients treated in our unit when compared to series reported in the past.33,34 The presence of lactate in the cerebrospinal fluid has been considered a marker of metabolic damage after brain injury that might impair postischaemic recovery.26,35,36 Brain lactic acidosis is a consequence of anaerobic metabolism and its predictive value for cerebral infarction and outcome has been suggested.32 On the other hand, the dissociations of CBF and metabolism, postulated as ‘luxury perfusion syndrome’ by Lassen in 1966,2 is associated with intracranial hypertension and might influence outcome since signs of ischaemia have been observed under this condition.37 In our study, 47 out of 50 patients (94%) had an abnormal AVDL that could be measured on one or more occasions (15.7% of AVDL measurements were abnormal). In order to identify ischaemia or hyperaemia patterns, the abnormal AVDL measurements were distributed in three different groups according to SjvO2 values that were measured simultaneously. We could not demonstrate any statistically significant association between the ‘luxury perfusion’ group (raised AVDL and SjvO2 ≥ 75%) and poor outcome. Finally, in order to elucidate the relationship between LOI and outcome in our series, we studied a subgroup of 33 patients who presented an LOI of 0.08 or greater in any of the periodical measurements done. Robertson et al. demonstrated that abnormal LOI are typical in ischaemia/infarction after brain injury;10 nevertheless, other causes for cerebral lactic acidosis, such as ventriculitis, could also cause an increase in the LOI in the absence of ischaemia. We found that in only 0.5% of all measurements could an abnormal LOI be calculated. Since most of the episodes occurred independently of each other and did not persist in subsequent measurements, no association with outcome could be demonstrated.
CONCLUSIONS This study includes a series of 50 patients with severe head injury accompanied in 42% by multiple injuries, who were rapidly resuscitated and immediately transferred to the hospital, certain features such as age, CT findings, CPP deterioration episodes and clinical severity (as judged by the GCS and pupil reaction, injury severity score and APACHE score), did not show statistically significant correlations with outcome. Only the occurrence of two or more episodes of jugular venous desaturation was found to correlate with the clinical outcome. The nature of the population studied and the relatively small number of patients might, nevertheless, influence our results.
© 2000 Harcourt Publishers Ltd
Cerebral oxygenation and outcome after brain injury 233
ACKNOWLEDGEMENTS The authors thank the physicians and nurses who work in our unit and have contributed to this study. We are also indebted to Mr Roland Stillhard for the preparation of graphical material.
19.
20. REFERENCES 1. Chan KH, Dearden NM, Miller JD, et al. Multimodality monitoring as a guide to treatment of intracranial hypertension after severe brain injury. Neurosurgery 1993; 32: 547–533. 2. Lassen NA. The luxury-perfusion syndrome and its possible relation to acute metabolic acidosis localised within the brain. Lancet 1966; 2: 1113–1115. 3. Miller JD, Becker DP, Ward JD, Sullivan HG, Adams WB, Rosner MJ. Significance of intracranial hypertension in severe head injury. J Neurosurg 1977; 47: 503–516. 4. Miller JD, Butterworth JF, Gudeman SK, Faulkner JE, Choi SC, Selhorst JB, Harbison JW, Lutz HA, Young HF, Becker DP. Further experience with the management of severe head injury. J Neurosurg 1981; 54: 289–299. 5. Muizelaar JP, Marmarou A, Ward JD, Kontos HA, Choi SC, Becker DP, Gruemer H, Young HF. Adverse effects of prolonged hyperventilation in patients with severe head injury: randomized clinical trial. J Neurosurg 1991; 75: 731–739. 6. Cruz J. Combined continuous monitoring of systemic and cerebral oxygenation in acute brain injury: preliminary observations. Crit Care Med 1993; 21: 1225–32. 7. Cruz J, Miner M, Allen S, Alves W, Gennarelli T. Continuous monitoring of cerebral oxygenation in acute brain injury: injection of mannitol during hyperventilation. J Neurosurg 1990; 73: 725–730. 8. Cruz J, Miner M, Allen S, Alves W, Gennarelli T. Continuous monitoring of cerebral oxygenation in acute brain injury: assessment of cerebral hemodynamic reserve. Neurosurgery 1991; 29: 743–749. 9. Cruz J, Raps E, Hoffstad O, Jaggi J, Gennarelli T. Cerebral oxygenation monitoring. Crit Care Med 1983; 21: 1242–1246. 10. Robertson CS, Narayan RK, Gokaslan ZL, Pahwa R, Grossman RG, Caram P Jr, Allen E. Cerebral arteriovenous oxygen difference as an estimate of cerebral blood flow in comatose patients. J Neurosurg 1989; 70: 222–230. 11. Sheinberg M, Kanter M, Robertson CS, Contant C, Narayan R, Grossman RG. Continuous monitoring of venous oxygen saturation in head-injured patients. J Neurosurg 1992; 76: 212–217. 12. Gopinath SP, Robertson CS, Contant CF, Hayes C, Feldman Z, Narayan RK, Grossman RG. Jugular venous desaturation and outcome after head injury. J Neurol Neurosurg Psychiatry 1994; 57: 717–723. 13. Valadka AB, Gospinath SP, Contant CF, Uzura M, Robertson CS. Relationship of brain PO2 to outcome after severe head injury. Crit Care Med 1998; 26: 1576–1581. 14. Baker SP, O’Neill B, Haddon W, Long WP. The injury severity score: a method for describing patients with multiple injuries and evaluating emergency care. J Trauma 1974; 14: 187–9. 15. Baker SP, O’Neill B. The injury severity score: an update. J Trauma 1976; 16: 882–5. 16. Knaus W, Draper E, Wagner, D, Zimmermann J. APACHE II: A severity of disease classification system. Crit Care Med 1985; 13: 818–829. 17. Teasdale G, Jennett B. Assessment of coma and impaired consciousness. Lancet 1974; 2: 81–84. 18. Stocker R., Bernays R., Kossmann T, Imhof HG. Monitoring and treatment of acute head injury. In: Goris E, Trentz O (eds). The integrated approach to
© 2000 Harcourt Publishers Ltd
21.
22.
23. 24.
25. 26.
27.
28.
29. 30.
31.
32.
33. 34. 35.
36.
37.
trauma care. The first 24 hours. Berlin, Heidelberg and New York: Springer Verlag, 1995; 196–210. Stocker R, Fandino J, Kossmann T, Trentz O, Imhof HG. Experiences in primary intensive care. In: Diemath HE, Sommerauer J, von Wild KRH (eds). Brain protection in severe head injury. Munchen, Bern, Wien and New York: W. Zuckshwerdt Verlag, 1996; 60–66. Marshall L, Marshall S, Klauber MR, Berkum Clark M, Eisenberg H, Jane J, Luerssen T, Marmarou A, Foulkes M. A new classification of head injury based on computerized tomography. J Neurosurg 1991; 75: S14–20. Andrews PJ, Dearden NM, Miller JD. Jugular bulb cannulation technique and validation of a new continuous monitor. Br Journal of Anaesthesia 1991; 67: 553–558. Jakobsen M, Enevoldsen E. Retrograde catheterization of the right internal jugular vein for serial measurements of cerebral venous oxygen content. J of Cereb Blood Flow and Metab 1989; 9: 717–720. Jennett B, Bond M. Assessment of outcome after severe brain damage. Lancet 1975; 1: 480–484. Marmarou A, Anderson RL, Ward JD, Choi SC, Young HF. Impact of ICP instability and hypotension on outcome in patients with severe head injury. J Neurosurg 1991; 75: 559–566. Adams JH, Graham DI, Gennarelli TA. Diffuse axonal injury in head injury caused by a fall. Lancet 1984; 22: 1420–1422. Kalimo H, Rehcrona S, Soderfeldt B, Olsson Y, Siesjö BK. Brain lactic acidosis and ischemic cell damage: 2. Histopathology. J Cereb Blood Flow Metab 1981; 1: 313–327. Siesjo BK, Katsura K, Zhao Q, Folbergrova J, Pahlmark K, Siesjo P, Smith ML. Mechanism of secondary brain damage in global and focal ischemia: a speculative synthesis. J Neurotrauma 1995; 12: 943–956. Krieger D, Adams H, Schwarz S, Rieke K, Aschoff A, Hacke W. Prognostic and clinical relevance of pupillary responses, intracranial pressure monitoring, and brainstem auditory evoked potentials in comatose patients with acute supratentorial mass lesions. Crit Care Med 1993; 21: 1944–1950. Ropper AH. Lateral displacement of the brain and level of consciousness in patients with acute hemispheral mass. N Engl J Med 1986; 314: 953–958. Eisenberg HM, Gary HE Jr, Aldrich EF, Saydjari C, Turner B, Foulkes MA, Jane JA, Marmarou A, Marshall LF, Young HF. Initial CT findings in 753 patients with severe head injury. J Neurosurg 1990; 73: 688–698. Coplin WM, O’Keefe GE, Grady MS, Grant GA, March KS, Winn HR, Lam AM. Thrombotic, infectious, and procedural complications of the jugular bulb catheter in the intensive care unit. Neurosurgery 1997; 41: 101–109. Robertson CS, Grossman R, Goodman C, Narayan R. The predictive value of cerebral anaerobic metabolism with cerebral infarction after head injury. J Neurosurg 1987; 67: 361–368. Dolder E. Fortschritte in der Behandlung von schweren Schaedel-HirnTrauma-Verletzung? Unfallchirurgie 1983; 94: 116–121. Zangger P, Bossert W, Heinz C. Bedürfnisanalyse und Bilanz 5 Jahre nach schweren Hirnverletzungen. Z. Unfallchirur Vers Med 1991; 84: 91–94. Rabow L, DeSalles AF, Becker DP, Yang M, Kontos HA, Ward JD, Moulton RJ, Clifton G, Gruemer HD, Muizelaar JP. CSF brain creatine kinase levels and lactic acidosis in severe head injury. J Neurosurg 1986; 65: 625–629. Rehcrona S, Rosen I, Siesjö BK. Brain acidosis and ischemic cell damage: 1. Biochemistry and neurophysiology. J of Cer Blood Flow and Met 1981; 1: 297–311. Obrist W, Langfitt T, Jaggi J, Cruz J, Gennarelli T.: Cerebral blood flow and metabolism in comatose patients with acute head injury. J Neurosurg 1984; 61: 241–253.
Journal of Clinical Neuroscience (2000) 7(3), 226–233
Clinical study
Cerebral oxygenation and systemic trauma related factors determining neurological outcome after brain injury Javier Fandino1 MD, Reto Stocker2 MD, Sonja Prokop2 RN, Otmar Trentz 2 MD, Hans-Georg Imhof1 MD 1
Department of Neurosurgery, Division of Trauma, University Hospital Zurich, Switzerland and 2Department of Surgery, Division of Trauma, University Hospital Zurich, Switzerland
Summary We examined the relationship between clinical and radiological findings, cerebral oxygenation patterns during intensive care management, presence of systemic trauma related injuries and severity of illness in 50 patients (age: 32.3 ± 12 years, GCS: 8 ± 4) who were rescued from the accident scene within a 30 min period after trauma. Presence of systemic injuries was quantified using the Injury Severity Score (ISS) and severity of illness was scored using the Acute Physiology and Chronic Health Evaluation (APACHE II). Cerebral oxygenation parameters included continuous monitoring of jugular bulb oxygen saturation (SjvO2) for 12 840 h, and 2323 periodical blood sampling for measurement of arteriovenous differences in oxygen content (AVDO2), arteriovenous difference of lactate (AVDL) and lactate oxygen index (LOI). Fifteen patients (30%) presented with anisocoria or non-reacting pupils. Diffuse lesions on computed tomography (CT) were found in 34% of the patients and in 66% a mass lesion was removed. The mean ISS was 28 ± 15.3 and 34 patients (68%) had an APACHE II score between 20 and 29 (mean 24 ± 15). No statistically significant association between age (P = 0.45), gender (P = 0.83), initial Glasgow Coma Score (GCS) (P = 0.43), episodes of cerebral perfusion pressure (CPP) < 70 mm Hg (P = 0.8), ISS (P = 0.28), pupillary abnormalities (P = 0.57), initial CT findings (P = 0.74), APACHE II scores (P = 0.36) and outcome could be demonstrated. The number of SjvO2 desaturations (< 60%) was the only statistically significant factor associated with outcome (P = 0.05). The percentage of patients with poor neurological outcomes (GOS 1–3) was 38% in patients with no or one desaturation episode, and 57.6% in those with multiple desaturations. In conclusion, in patients who are resuscitated early and quickly transferred to the hospital, the number of SjvO2 desaturations during intensive care management might be associated with outcome more strongly than other clinical and radiological features. © 2000 Harcourt Publishers Ltd Keywords: severe head injury, Injury Severity Score, APACHE II, intensive care, jugular bulb oxygen saturation, outcome
INTRODUCTION
Table 1 Demographic and trauma-related factors
Several studies have demonstrated that physiological factors like systemic hypotension, intracranial hypertension, arterial hypoxia and hypocapnia are associated with poor outcome after brain injury.1–5 Introducing new techniques into clinical practice, such as jugular bulb oxygen saturation (SjvO2), allowed a better understanding of cerebral oxygenation patterns during intensive care management of head injured patients.6–9,10,11 The incidence of SjvO2 desaturation episodes and, more recently, low brain tissue oxygen pressure values (PbtO2 below 15 torr), were reported to be associated with a poor neurological outcome.12,13 Brain injury must be considered and managed as a systemic condition in which many extracranial factors may influence outcome. Besides clinical and radiological findings, well established scores of systemic injuries, such as the Injury Severity Score (ISS), and scores denoting the severity of systemic illness, like the Acute Physiology and Chronic Health Evaluation score (APACHE II), have been routinely used for assessing the prognosis of trauma patients who receive intensive care management.14–16 The purpose of the present study was to examine the relationship between clinical and radiological findings, systemic trauma related factors such as ISS and APACHE II, cerebral oxygenation parameters, and outcome in severe head injured patients who were resuscitated within a 30 min period after injury and rapidly transferred to the hospital.
Total of patients Age (mean ± SD) Sex (ratio M:F) Male (%) Female (%) Glasgow Coma Scale score (mean ± SD) 3–8 (%) 9–12 (%) 13–15 (%) Trauma mechanism Fall Car Motorcycle Assault/gunshot Sport Bicycle Others ISS (mean ± SD) Type of injury Open/isolated (%) Open/multitraumatised (%) Closed/isolated (%) Closed/multitraumatised (%) APACHE II (mean ± SD) Duration of ICU management (mean ± SD) Glasgow Outcome Scale score (mean ± SD)* D (%) VS and SD (%) MD and GR (%)
Received 23 February 1999 Accepted 25 May 1999 Correspondence to: Javier Fandino MD, Department of Neurosurgery, University Hospital Zurich, Frauenklinkstrasse 24, Zurich, Switzerland. Tel: +41–1–2552660; Fax: +41–1–2554505; E-mail: [email protected]
226
50 32.3 ± 12 years 2.3 35 (70) 15 (30) 8±4 26 (52) 14 (28) 10 (20) 16 (38%) 10 (22%) 7 (14%) 5 (10%) 2 (4%) 2 (4%) 7 (14%) 28 ± 15.3 24 (48) 1 (2) 5 (10) 20 (40) 24 ± 5 28.4 ± 15.2 days 4±1 9 (18) 16 (32) 25 (50)
Abbreviations: SD: standard deviation; ISS: Injury Severity Score;3,4 APACHE: Acute Physiology and Chronic Health Evaluation;17 ICU: Intensive Care Unit; *: Outcome 3 months after trauma (D: death, VS: vegetative state; SD: severe disability; MD: moderate disability; GR: good recovery).
Cerebral oxygenation and outcome after brain injury 227
CLINICAL MATERIAL AND METHODS Patient population, clinical assessment and management Fifty consecutive patients with severe brain injuries admitted to the Trauma Intensive Care Unit (TICU) during a 32 month period were included in this study. The demographic characteristics are shown in Table 1. As entry criteria patients had to be rescued and assessed in the scene of the accident within the first 30 min after injury, and immediately transferred to our hospital by helicopter or ambulance equipped with mechanical ventilator and intensive care equipment. The mean age was 32.3 ± 12 years (range 17–60). Seventy percent of the patients were men (n: 35) and 30% were women (n: 15). The mechanism and cause of trauma was verified using the resuscitation and referral reports. Eighteen patients (40%) suffered head injury in transport related accidents (car, motorcycle and bicycle). None of the bicycle riders (4%) was using a helmet. A fall was the trauma mechanism in 16 patients (32%). Violence related episodes were the cause of head injuries in 5 patients (10%). Sports related injuries were found in 2 patients (4%) and other causes in 2 patients (4%). Patients with fatal gunshot injuries or brain death suspicion on admission were not included in the study. Twenty-five patients (50%) had open head injuries, most of whom (48%) had isolated brain injuries. On the other hand, 25 patients (50%) suffered closed brain injuries, the majority of which were multitraumatised (40%). A total of 29 patients (58%) presented with isolated brain injuries and 21 patients (42%) were multitraumatised. Glasgow Coma Score17 (GCS) was assessed at the scene of the accident and on admission to the hospital. Considering the aim of this study and the possible multiple factors influencing the outcome, we did not define severity of brain injury in terms of GCS. Correlation of GCS and outcome was analysed as a score and not as a category of severity of brain injury. The majority of the patients (n: 38, 76%) were intubated at the scene of the accident after neurological evaluation and came to the emergency room paralysed and sedated. Pupillary size and response were documented at the scene of the accident, emergency room and periodically after admission in the TICU. Oxygen saturation (SaO2) was immediately monitored in the rescue scene and, when possible, blood gas analyses were made in order to rule out a hypoxic episode prior to resuscitation or admission into the hospital. All patients were classified according to the ISS.14,15 After admission in the TICU, patients were assigned specific diagnostic categories and scored using the APACHE II which provides a general measure of severity of disease.16 APACHE II uses a point score based upon initial values of 12 routine physiologic measurements, age and previous health status. At TICU admission, any emergency operation was recorded for later mortality risk prediction. Our management protocol has been described elsewhere.18,19 Therapeutic tools included circulatory monitoring with a pulmonary artery catheter (Swan-Ganz), invasive mean arterial pressure monitoring (MAP), cerebrospinal fluid (CSF) drainage, hyperventilation under SjvO2 monitoring, osmotherapy with mannitol and mild hypothermia. Sixteen patients (32%) underwent barbiturate coma the thiopentone dose was adjusted according to bedside; electroencephalography (EEG) registration (6 bursts/min). Corticosteroids were not included in our treatment protocol and no antiepileptic drugs were administrated phrophylactically. Continuous intracranial pressure (ICP) monitoring was performed on all patients. Induced hypertension with cathecolamines and slight hypervolaemia was maintained in order to keep cerebral perfusion pressure (CPP) over 70 mm Hg. The most common ICP device used was a subdural catheter (49%). Ventriculostomy was performed in 32% of the patients. In 2% of the patients ICP was © 2000 Harcourt Publishers Ltd
monitored with infrared devices (Camino Laboratories,® San Diego CA). A wake up procedure was indicated if no ICP increase above 15 mm Hg within 24 h happened under normothermia and normoventilation, if the amount of CSF drainage was below 80 ml/24 h, if SjvO2 and AVDL values were normal, and if no signs of intracranial hypertension was evident on the computed tomography (CT) scan. ICP monitoring was discontinued as soon as neurological assessment was possible and no therapeutic interventions were necessary for 24 h. Radiological evaluation All patients underwent radiological evaluation with skull projections and CT immediately after admission to the hospital. Initial CT scans were evaluated according to the Traumatic Coma Data Bank (TCDB):20 categories 1–4 (diffuse injuries); 5 (evacuated mass lesion); and 6 (non-evacuated lesion). During ICU management, patients underwent CT follow up within 48 h after admission, before the wake up procedure if ICP increased or became intractable with critical CPP, or if unexplained deterioration of cerebral oxygenation parameters appeared. Cerebral angiography was performed in patients with presence of atypical posttraumatic subarachnoid haemorrhage (tSAH) and suspicion of ruptured aneurysm, traumatic arteriovenous fistulae (cervical, dural, cortical or carotid-cavernous) or venous sinus injuries. Bedside SjvO2 monitoring and measurements of AVDL To monitor SjvO2, a No. 5.5 French fiberoptic oxygen double lumen catheter was inserted percutaneously into the internal jugular vein through a No. 6 French introducer sheath according to the technique describe in detail elsewhere.21,22 The position of the catheter was controlled radiologically with anteroposterior and lateral skull projections. The catheter for measurement SjvO2 was placed on the right side unless a unique unilateral lesion in the left hemisphere was present, the left jugular circulation was demonstrated to be dominant, or if thrombosis or thrombi in the right jugular vein were documented in the colour coded duplex sonography. On-line printouts of SjvO2 curves were available. Every effort was made to start the monitoring as early after admission as possible. Pre-insertion and in vivo calibration was performed after insertion and every 6 h thereafter. If SjvO2 changed abruptly or if technical measurement problems were suspected, in vivo calibration was immediately repeated and documented. SaO2 and haemoglobin concentration was, if possible, kept constant throughout the ICU management, depending on presence of lung or chest injuries. Normal SjvO2 values were considered between 60% and 75%. We defined a desaturation episode as SjvO2 lower than 60%. Causes of artifactual measurement were ruled out before considering cerebral hypoperfusion using an algorithm described elsewhere.10 SjvO2 greater than 75% suggested luxury perfusion.19 Two patient groups were considered for the analysis of outcome: patients with no or one desaturation episode, and patients with multiple desaturation episodes. If desaturation was suspected based on the fiberoptic measurements, blood gas analysis was done to confirm the values. AVDO2 was calculated from the difference between arterial and jugular vein oxygen content. Arteriovenous difference of lactate (AVDL) was calculated subtracting jugular vein from arterial (systemic) lactate concentration values. AVDL and SjvO2 (and AVDO2) were measured in order to identify simultaneous episodes of luxury perfusion and ischaemia. AVDL values of 0.2 µmol/L or greater were interpreted as increased cerebral lactate production due to an ischaemia/infarction episode. Lactate oxygen index LOI (LOI = –AVDL/AVDO2) was calculated according to Robertson et al.10 A LOI of 0.08 or greater was interpreted as an Journal of Clinical Neuroscience (2000) 7(3), 226–233
228 Fandino et al. Table 2 Demographic characteristics and outcome
Mean age (SD) M:F ratio (% men) Mean initial GCS (SD) Mean Injury Severity Score (SD) Mean APACHE II (SD) Initial CT findings (TCDB classification) Patients with diffuse injuries: categories 1–4 (%) Patients with mass lesions: categories 5–6 (%) Patients with pupillary abnormalities (%) Patients with CPP < 70 mm Hg (%) No episode One episode Two or more episodes Number of SjvO2 desaturation episodes/Percentage of patients who presented SjvO2 desaturation episodes Number of patients (%)
GOS 4–5
GOS 2–3
GOS 1
P-value*
30.5 (12.8) 7.3 (88) 10 (3.4) 25.6 (5.8) 21.1 (3.8)
34.7 (12.2) 1.6 (62) 7 (3.8) 29.7 (7.2) 25.2 (2.2)
33.2 (13.4) 0.5 (33) 7 (3.6) 31.4 (4.4) 24.8 (5.6)
0.45 0.83 0.46 0.28 0.36 0.74
9 (18) 16 (34) 8 (16)
5 (10) 11 (20) 4 (8)
3 (6) 6 (12) 3 (6)
11 (22) 6 (12) 8 (16) 93/48
10 (20) 1 (2) 5 (10) 134/56
5 (10) 2 (4) 2 (4) 11/71
25 (50)
16 (32)
9 (18)
0.57 0.8
0.05
Abbreviations: GOS: Glasgow Outcome Scale score 3 months after trauma;15 SD: standard deviation; M: male; F: female; GCS: Glasgow Coma Scale score; APACHE: Acute Physiology and Chronic Health Evaluation;17 CT: computed tomography; TCDB: National Institutes of Health Traumatic Coma Data Bank;20 SjvO2: jugular bulb oxygen saturation; *Significance considering GOS groups.
ischaemia/infarction pattern. All parameters (ICP, CPP, mean arterial pressure (MAP), SjvO2, AVDO2, AVDL, SaO2, SvO2, PaO2 and PCO2) were documented in a ‘cerebral metabolism’ database at least every 6 h. Lactate measurements were done in the laboratory (TDx FLx lactate analyzer, Abbott Laboratories, Abbott Park, IL, USA). Blood gas analyses were performed in the unit (ABL System, Radiometer, Radiometer Medical A/S, Copenhagen, Denmark).
patients (44%) had GR, 10 patients (20%) were MD and 6 patients (12%) were SD. There was no statistically significant difference in age (P = 0.45) or gender (P = 0.83) distribution in relation to GOS (Table 2). Patients who had GOS 4 or 5 (GD or MD) were in 84% of the cases males and were 30.5 years (SD 12.8). Patients with a GOS of 2 or 3 (MD or SD) were in 62% of the cases males (m:f ratio = 1.6) with an age of 34.7 ± 12.2 years. Finally, 33% of patients who died (GOS 1) were males (m:f ratio = 0.5) with an age of 33.2 ± 13.4 years.
Determination of outcome The Glasgow Outcome Score (GOS) was used to quantify the severity of the neurological deficits.23 Outcome was determined 3 months and one year after injury by personnel who were unaware of the cerebral oxygenation data obtained during the intensive care management. The patients were routinely controlled at the neurological department in a special out-patient unit for brain injured patients. A standard battery of memory tests, social integration assessment and neurological examination were included in the evaluation. Statistical analysis Summary data are reported as the mean ± standard deviation. Categorical variables of assessment of brain injury and jugular vein saturation values were analysed using Chi-square whereas differences in mean were tested using a one-way analysis of variance. The significant level used to determined statistical significance was 0.05. RESULTS Outcome after three months and one year after trauma Outcome evaluation 3 months after brain injury showed that 8 patients (16%) had good recovery (GR), 18 patients (36%) were moderately disabled (MD), 14 patients (28%) were severely disabled (SD), 1 patient (2%) was in a vegetative state (VS) and 9 patients (18%) died (D) (Tables 1 and 2). Six patients died because of intractable intracranial hypertension and three because of medical complications associated with deterioration of CPP. One patient, who was in VS, died eight months after discharge from the hospital because of pulmonary complications. GOS, after one year follow up, showed an improved outcome in 20 patients (40%): 22
Journal of Clinical Neuroscience (2000) 7(3), 226–233
Initial clinical and neurological assessment The average of initial GCS was 8 ± 4. Twenty-six patients (52%) presented with initial GCS ≤ 8, 14 patients (28%) had GCS between 9 and 12, and 10 patients (20%) had GCS between 13 and 15. As shown in Table 2, patients with an initial GOS of 4 or 5 (GR or MD) had GCS of 10.3 ± 3.4. Patients with lower GOS (SD, VS or D) had a similar initial GCS. Patients with GOS 2 or 3 (SD or VS) had an initial GCS of 7.6 ± 3.8, and patients who died (GOS 1) had a similar GCS (7.4 ± 3.6). In our series, initial GCS was not significantly significant associated with outcome (P = 0.46). Initial pupillary examination revealed that 70% (n: 35) of the patients had normal and reacting pupils. Thirty percent of the patients (n: 15) presented with anisocoria or non-reacting pupils. We found no statistically significant relationship between pupillary abnormalities and outcome after 3 months (P = 0.34). At least in our series, there were more patients who presented with anisocoria and had GOS 4 or 5 (n: 8) than patients who had GOS 2 or 3 (n: 4) or died (n: 3). Initial computed tomography and outcome Diffuse lesions (TCDB Categories 1–4) were found in 17 patients (34%) and mass lesions (TCDB Categories 5–6) were present in 33 patients (66%). Nineteen patients (38%) underwent craniotomy and removal of a haematoma immediately after admission. The number of patients who presented with mass lesions was twice as many as patients who had diffuse injury of the brain in all GOS groups (Table 2). Sixteen patients (32%) with mass lesions, compared to 9 patients (18%), had a GOS 4 or 5 after 3 months. Eleven patients (22%) with initial mass lesions on CT, compared to 5 patients (10%), were severely disabled or persistently vegetative in the outcome evaluation. In the group of
© 2000 Harcourt Publishers Ltd
Cerebral oxygenation and outcome after brain injury 229
Fig. 1 Injury Severity Score (ISS)2,3 and GOS 3 months after trauma. No statistically significant correlation could be demonstrated (P = 0.28). GOS: 1: death; 2: vegetative state; 3: severe disabled; 4: moderate disabled; 5: good recovery.
Fig. 3 Acute Physiology and Chronic Health Evaluation (APACHE II17) and GOS three months after trauma. No statistically significant correlation could be demonstrated (P = 0.36). GOS: 1: death; 2: vegetative state; 3: severe disabled; 4: moderate disabled; 5: good recovery.
lower than 10. There were 3 patients in each APACHE II category group. The mean APACHE II in patients with GOS 4 and 5 was 21 ± 3.8, in patients with GOS 2 and in patients who died the mean was 24.8 ± 5.6. APACHE II was not a mortality predictor in this series and, as shown in Figure 3, the relationship between APACHE II and GOS was not statistical significant (P = 0.36). CPP deterioration and neurological outcome
Fig. 2 Relation of Acute Physiology and Chronic Health Evaluation (APACHE II17) and mortality after brain injury.
patients who died, 6 had mass lesions (12%) and 3 (6%) presented with diffuse injuries. In this series, initial CT findings did not correlate significantly with outcome (P = 0.74). Injury severity score and outcome The mean initial ISS was 28 ± 15. The majority of the patients had an ISS between 20–39. Twenty-eight patients (56%) presented with an ISS between 20 and 29, and 10 patients (20%) had an ISS between 30 and 39. Only 5 patients (10%) had an ISS greater than 40. Thirty-eight percent of the patients included in this study underwent extracranial emergency operation under ICP monitoring to ensure adequate CPP intraoperatively. Figure 1 shows the relation between ISS and the outcome 3 months after injury. Patients with GOS 4 or 5 had an ISS of 25 ± 6, patients with GOS 2 or 3 had a ISS of 29 ± 7, and patients who died had an ISS of 31 ± 4. In the group of ISS < 20 (n: 7), the majority of the patients had a GOS 2 or 3 (30%). The majority of patients who died (GOS 1) had an ISS between 30 and 39. The association between ISS and GOS was not statistically significant in this series (P = 0.28). Acute physiology and chronic health evaluation (APACHE II) We classified our patients into three APACHE score category groups (Fig. 2). Thirty-four patients (68%) had a score between 20 and 29, 11 patients (22%) had a score between 10 and 19, and 5 patients had an APACHE of more than 30. No patient had a score © 2000 Harcourt Publishers Ltd
Table 2 summarises the number of patients presenting episodes of CPP less than 70 mm Hg. In a total of 26 patients (52%) CPP could be maintained over 70 mm Hg, 9 patients (18%) one episode of CPP deterioration which required therapeutic intervention could be documented. Finally, 15 patients (30%) had two or more persistent episodes of CPP > 70 mm Hg which required any therapeutic intervention. No statistically significant association could be demonstrated between CPP deterioration and neurological outcome. SjvO2 desaturation episodes and neurological outcome A total of 12 840 h of on-line SjvO2 monitoring during mechanical ventilation were collected. In addition 2323 measurements of AVDO2 and AVDL at 6 h intervals were performed. The mean duration of intensive care management in this series was 28.4 ± 15.2 days. The long duration of ICU care was due mainly to the fact that 32% of the patients underwent barbiturate coma. The duration of continuous SjvO2 monitoring ranged between 2 and 35 days (mean, 12 ± 3 days). All patients underwent jugular vein cannulation for SjvO2 monitoring and AVDL measurements before starting hyperventilation. In our series, hyperventilation without signs of anaerobic metabolism due to ischaemia/infarction signs (AVDL ≥ 0.2 µmol/L and/or SjvO2 < 60%) was possible in only 13 of 50 patients (26%). In 37 patients (74%) hyperventilation was contraindicated due the risk of ischaemia. A total of 238 SjvO2 desaturation episodes were detected or confirmed in the periodical jugular bulb blood sampling (10.2% from the total of periodical measurements). Desaturation episodes were most common within the first 48 h after injury (68%). The incidence of desaturation and percentage of patients who presented one or more desaturation episodes was higher in patients with poor outcomes (Table 2). A higher percentage of desaturations was observed among the patients who died (71%), compared to those patients who had GOS 2 or 3 (56%) or patients with GOS 4-or 5 (48%). As presented in Figure 4, the percentage of patients with poor neurological outcomes (GOS 1–3) was 38% in patients with no or one desaturation episodes, and 57.6% in those Journal of Clinical Neuroscience (2000) 7(3), 226–233
230 Fandino et al.
patients (66%). Seventeen patients did not show abnormal LOI during the intensive management. The majority of the patients had five or less episodes (measurements) of abnormal LOI (27 patients). Five patients presented more than five episodes of LOI of 0.08 or greater. No statistically significant association with outcome was found in patients with more than 5 episodes of LOI ≥ 0.08 (GOS 1–3: 1 patient, 4–5; 6 patients) or those with five or less episodes of abnormal LOI (GOS 1–3: 12 patients, 4–5: 14 patients). Complications of jugular vein catheterisation
Fig. 4 Correlation between SjvO2 desaturation episodes and GOS three months after trauma (P = 0.05). GOS: 1: death; 2: vegetative state; 3: severe disabled; 4: moderate disabled; 5: good recovery.
with multiple (2 or more) desaturation episodes. The percentage of patients with favourable outcomes (GOS 4–5) with no or one desaturation episode was 60%, compared to those with multiple desaturations (41.8%). Mortality in the group of patients who experienced multiple desaturations was higher (6 of 9 patients who died during intensive care management had multiple desaturations) than in patients who had no or one desaturation. The incidence of jugular venous desaturation episodes (none or one compared to multiple) was the only factor associated with neurological outcome in this study (P = 0.05). Simultaneous SjvO2 and AVDL abnormalities
The incidence of carotid puncture could not be clearly quantified retrospectively in our series; nevertheless, the introduction of Doppler and colour coded duplex guided punctures diminished the incidence of this complication throughout the study period. Subcutaneous haematoma after carotid puncture did not cause any major complications such as airway obstruction, diminished flow in the carotid artery or acute ICP elevations. The tips of the 105 catheters used in the 50 patients were cultured and no bacteriological growth could be demonstrated in any specimen. As soon as the puncture site showed inflammatory signs (started to redden, or serious fluid was detected) the catheter and introducer were replaced. Thrombosis of the internal jugular vein was documented only in a young female patient (Fig. 5). This woman died on day 12 after brain injury suffering intractable generalised brain oedema with uncontrollable electrolyte disorders associated with a severe coagulopathy. An immune complex disorder was diagnosed (positive Raji cell assay) and was considered to be a clinical factor associated with thrombus formation. DISCUSSION
In 47 patients (94%) abnormal lactate values (AVDL ≥ 0.2 µmol/L) were measured. A total of 366 AVDL measurements (15.7%) were found to be abnormally high as a consequence of brain lactic acidosis. Only three patients had normal AVDL throughout the intensive care management. Table 3 shows SjvO2 values in relation to GOS in patients with AVDL ≥ 0.2 µmol/L. The simultaneous measurement of SjvO2 in the 47 patients who presented one or more episodes of AVDL ≥ 0.2 µmol/L was 70% ± 9. Thirty-three patients (66%) presented luxury perfusion signs by means of abnormal AVDL and SjvO2 ≥ 75%; nevertheless, no significant association with poor outcome could be demonstrated (11 patients had GOS 1–3 and 22 patients had GOS 4–5). All patients who presented one or more episode of brain lactic acidosis during the course of the intensive care management also had one or multiple SjvO2 desaturations. No statistically significant differences were found in the analyses of the associations between the different SvjO2 subgroups and outcome in patients with abnormal AVDL. Lactate oxygen index A lactate oxygen index (LOI = –AVDL/AVDO2) of 0.08 or greater was found in 110 of 2326 measurements (0.5%) in 33 out of the 50
Initial clinical and radiological findings Predictive factors after brain injury have been extensively studied in the last two decades.3–5,24 In the early 1980s several studies led to the consensus that factors predicting poor outcome after brain injury include the presence of an intracranial haematoma, increasing age, abnormal motor responses, impaired or absent eye movements or pupil light reflexes, early hypotension, hypoxaemia or hypercarbia, and elevation of ICP over 20 mm Hg despite artificial ventilation.4 In addition, other studies reported poor outcome in patients with ICP instability and hypotension,20 prolonged hyperventilation5 and CPP under 70 mm Hg.1 Based on the experience gained in the last two decades, specialists in intensive care, paramedics and other disciplines other than neurosurgery have assumed a crucial role preventing secondary brain injuries. Diffusion of concepts and protocols in the medical and paramedical community, and the introduction of new monitoring techniques, have been important accomplishments in our country over the last decade.18,19 Switzerland, with an area of 41.293 sq. km, has traditionally been a pioneer in terrestrial and air patient transportation. Since most of our patients are rescued and transported immediately after trauma, we decided to analyse demographic, clinical
Table 3 Outcome in patients who presented one or more episode of brain lactic acidosis (AVDL ≥ 0.2 µmol/L) distributed according to SjvO2 values (P = NS). SjvO2 Group 61%–74% ≥ 75% ≤ 60%
No. measurements*
No. patients
Mean SjvO2
GOS 1–3
GOS 4–5
366 (15.7%) 110 (4.7%) 122 (5.2%)
47 (94%) 33 (66%) 47 (94%)
70 ± 9% 79 ± 6% 56 ± 6%
19 (40.5%) 11 (34.4%) 19 (40.5%)
28 (59.5%) 22 (66.6%) 28 (59.5%)
Abbreviations: GOS:15 Glasgow Outcome Score (percentage of number of patients in each SjvO2 group); *in parenthesis: percentage of total number of measurements from the total of 2326 periodical measurements done in 50 patients.
Journal of Clinical Neuroscience (2000) 7(3), 226–233
© 2000 Harcourt Publishers Ltd
Cerebral oxygenation and outcome after brain injury 231
Fig. 5 Longitudinal sonographic view of an obstructing thrombus (arrow head) in the internal jugular vein (ijv). The flow cannot be visualised beyond the entrance of the catheter in the vein (arrows).
and radiological factors influencing outcome in 50 consecutive patients resuscitated within a 30 min period after injury. Miller et al. published a study describing the factors important in predicting outcome based on 225 consecutive severely head injured patients.4 The authors reported that approximately half of the patients were referred from other hospitals and 20% were seen by a neurosurgeon within the first hour after injury. The understanding of pathophysiological mechanisms in diffuse axonal injury and cerebral ischaemia helped introduce the concept of the concept of secondary brain damage.25–27 Since then, efforts have made in this decade to emphasise the importance of early resuscitation, surgical treatment and intensive care management. Undoubtedly, outcome predictors have changed, especially in patients who are transferred to the hospital in an ultra-short period of time. In our series, age did not significantly influence outcome; nevertheless, we should mention that our patients’ ages were homogeneous and did not exceed 60 years (32 ± 12 years). Severity of initial radiological findings and initial GCS did not significantly differ in the three outcome groups (P = 0.46); nevertheless, one must consider that most of our patients are already intubated and mechanically ventilated at the scene of the accident, which does not allow for additional reliable clinical assessments. Patients who do not obey commands, or who have a GCS of 8 or less, were usually sedated and paralysed before admission. This policy might influence the predictive value of GCS; nevertheless, it enables better haemodynamic and respiratory control. Another clinical factor, which has been described as an outcome predictor in severely head injured patients is the pupil light response.4,28 According to Ropper, it is a horizontal shifting that gives rise to ipsilateral pupillary abnormalities by overstretching the peripheral pupillary pathways within the oculomotor nerve and does not result, as was previously thought, from the compression caused by downward shifting.29 Pupillary abnormalities were found to have prognostic value in a study with 55 comatose individuals with supratentorial mass lesions, including 15 patients © 2000 Harcourt Publishers Ltd
with brain injury.28 In the Richmond series, however, correlations between bilateral absence of the pupil response to light, impaired or absent oculocephalic responses and a poor outcome were present in both surgical (P < 0.001) and non-surgical patients (P < 0.001). In our series, 8 out of the 15 patients who presented with anisocoria survived having a GOS 4 or 5. Only 3 patients who presented with anisocoria died. In this study, we found no statistically significant relationship between pupillary abnormalities and outcome after 3 months (P = 0.34). Based on a prospective study of over 750 initial CT scans in patients with severe head injury, The National Institutes of Health Traumatic Coma Data reported that the status of the mesencephalic cisterns and the degree of midline shift are strongly associated with high ICP and death.30 In our study, in spite of the fact that no statistically significant differences were found when comparing the different outcome groups in relation to the initial CT classification, the presence of a mass lesion seemed to be an ominous predictor of poor outcome. Thirty-three patients with mass lesions died, were vegetative or were severely disabled compared to 8 patients in whom diffuse injuries were initially diagnosed. Finally, we have to highlight that our results are based exclusively on the evaluation of CT findings in terms of TCDB classification. The fact that other factors such as midline shift, compression of basal cisterns and presence of subarachnoid blood were not independently evaluated might constitute a limitation of this study. Severity of systemic illness: ISS and APACHE Considering that brain injury must be approached as a systemic condition, we analysed the ISS and APACHE along with the initial clinical and radiological findings in order to quantify the degree of systemic lesions and the patient’s systemic condition. ISS is a method for numerically describing the overall severity of injury;15 it can be applied to persons who have sustained injury to more than one area of the body as well as to those with isolated injuries. The Journal of Clinical Neuroscience (2000) 7(3), 226–233
232 Fandino et al.
ISS correlates substantially better with mortality than does the Abbreviated Injury Scale (AIS) rating for the single most severe injury.14 In spite of the fact that ISS was introduced in the early 1970s, it has not been used routinely in the evaluation of brain injured patients. Since multitrauma is frequent in patients with brain injury (42% of the patients in our series were multitraumatised and 58% presented with isolated brain injuries), we believe that this score is useful not only in the assessment of systemic injuries, but also as a factor which might influence outcome. Miller et al. reported that despite the association between multiple injury and some systemic insults, there was no significant relationship between the presence of multiple injuries and outcome.4 More recently, Gospinath et al. classified a series of patients using the ISS, but found no statistically significant relationship between ISS and outcome after brain injury.12 The ISS ‘lethal dose or score’ for 50% of the patients is 40 for ages 15–44, 29 for ages 45–64 and 20 for ages 65 or older.15 The mean ISS in our study was 28 ± 15.3. In patients who died, the ISS was 31.4 ± 4.4 and in patients who survived with a GOS 4 or 5, the ISS was lower (25.6 ± 5.8). No statistically significant difference, nevertheless, could be observed. According to the Baltimore and Birmingham series, a mortality rate of approximately 20% would be expected in our series, which might coincide with the mortality rate in our study (18%). In addition, ISS has proven to be exponentially related to the grade of disability after trauma. Patients with an ISS of approximately 18 were reported to be severely disabled after trauma; in our study, patients severely disabled (GOS 2 or 3) presented with a much higher ISS (29.7 ± 7.2). These observations suggest that ISS might predict mortality but not disability in brain injured patients. The basis for the development of APACHE was the hypothesis that the severity of acute disease can be measured by quantifying the degree of abnormality from multiple physiologic variables.16 The number of physiological measurements in the original APACHE introduced in 1981 have been reduced from 34 to 12. The new version of APACHE or APACHE II include the following parameters: temperature, MAP, heart rate, respiratory rate, oxygenation measurements, arterial pH, serum sodium, serum potassium, serum creatinine, haematocrit, white blood count, GCS, age and a so called chronic health point for emergency and elective postoperative patients. In a multicentric validation study, Knaus et al. evaluated APACHE II in 5815 ICU admissions and demonstrated a direct relationship between high scores and the hospital death rates.16 For each five point increase in APACHE II, there was a significant increase in death rate; nevertheless, the overall risk of hospital death varied according to the disease. In addition, non-operative and postoperative patients have different death rates in the different APACHE II groups. In our study, the mean APACHE II score was 24 ± 5. According to APACHE II, a mortality rate between 30% and 40% would be expected in our series (in both operative and non-operative patients); nevertheless, a lower hospital death rate of 18% was, in fact, observed. On the other hand, APACHE II scores did not differ significantly in the three outcome groups (GOS 1: 24.8 ± 5.6, GOS 2–3: 25.2 ± 2.2 and GOS 4–5: 21.1 ± 3.8). One speculative explanation of the discordance between the expected and observed mortality rates could be the diversity of diagnoses included in the APACHE II study. Among the patients included in the study, only 2.3% had suffered head injury (both non-operative and operative) and only 4.8% were admitted because of multiple trauma. Our results might suggest that APACHE II is not a reliable predictor of hospital death in head injured patients, even if they presented as multitraumatised. Regardless of this fact, as the authors of the APACHE study themselves admit, prognostic estimates are still only estimates and devotion of the ICU personnel in the care of patients will always play the most important role in outcome.
Journal of Clinical Neuroscience (2000) 7(3), 226–233
Cerebral oxygenation patterns and outcome Since its introduction in the early 1980s, SjvO2 monitoring has become an important tool for understanding oxygenation patterns after brain injury.9 Complications resulting from this technique have been reported to be few and minor.31 SjvO2 enables the diagnosis of jugular vein desaturation episodes, brain lactic acidosis and reliable estimates of cerebral blood flow (CBF).10,11,32 The occurrence of jugular venous desaturation has been shown to be associated with poor neurological outcome.12 In our study we demonstrated that the percentage of patients who presented desaturation episodes was higher in patients who died (71%) than in patients who had GOS 2–3 (56%) or GOS 4–5 (48%). In addition, the percentage of patients with favourable outcomes who had no or one desaturation episode was significantly higher than among those who had multiple episodes. These observations, which highlight the importance of assessing the global cerebral oxygen supply and demand as an outcome predictor, were recently reinforced by studying the relationship between focal brain tissue oxygen pressure (PbtO2) and outcome after severe head injury.13 We believe that understanding cerebral oxygenation patterns and quickly detecting hypoperfusion states have led to better outcome rates in severely head injured patients treated in our unit when compared to series reported in the past.33,34 The presence of lactate in the cerebrospinal fluid has been considered a marker of metabolic damage after brain injury that might impair postischaemic recovery.26,35,36 Brain lactic acidosis is a consequence of anaerobic metabolism and its predictive value for cerebral infarction and outcome has been suggested.32 On the other hand, the dissociations of CBF and metabolism, postulated as ‘luxury perfusion syndrome’ by Lassen in 1966,2 is associated with intracranial hypertension and might influence outcome since signs of ischaemia have been observed under this condition.37 In our study, 47 out of 50 patients (94%) had an abnormal AVDL that could be measured on one or more occasions (15.7% of AVDL measurements were abnormal). In order to identify ischaemia or hyperaemia patterns, the abnormal AVDL measurements were distributed in three different groups according to SjvO2 values that were measured simultaneously. We could not demonstrate any statistically significant association between the ‘luxury perfusion’ group (raised AVDL and SjvO2 ≥ 75%) and poor outcome. Finally, in order to elucidate the relationship between LOI and outcome in our series, we studied a subgroup of 33 patients who presented an LOI of 0.08 or greater in any of the periodical measurements done. Robertson et al. demonstrated that abnormal LOI are typical in ischaemia/infarction after brain injury;10 nevertheless, other causes for cerebral lactic acidosis, such as ventriculitis, could also cause an increase in the LOI in the absence of ischaemia. We found that in only 0.5% of all measurements could an abnormal LOI be calculated. Since most of the episodes occurred independently of each other and did not persist in subsequent measurements, no association with outcome could be demonstrated.
CONCLUSIONS This study includes a series of 50 patients with severe head injury accompanied in 42% by multiple injuries, who were rapidly resuscitated and immediately transferred to the hospital, certain features such as age, CT findings, CPP deterioration episodes and clinical severity (as judged by the GCS and pupil reaction, injury severity score and APACHE score), did not show statistically significant correlations with outcome. Only the occurrence of two or more episodes of jugular venous desaturation was found to correlate with the clinical outcome. The nature of the population studied and the relatively small number of patients might, nevertheless, influence our results.
© 2000 Harcourt Publishers Ltd
Cerebral oxygenation and outcome after brain injury 233
ACKNOWLEDGEMENTS The authors thank the physicians and nurses who work in our unit and have contributed to this study. We are also indebted to Mr Roland Stillhard for the preparation of graphical material.
19.
20. REFERENCES 1. Chan KH, Dearden NM, Miller JD, et al. Multimodality monitoring as a guide to treatment of intracranial hypertension after severe brain injury. Neurosurgery 1993; 32: 547–533. 2. Lassen NA. The luxury-perfusion syndrome and its possible relation to acute metabolic acidosis localised within the brain. Lancet 1966; 2: 1113–1115. 3. Miller JD, Becker DP, Ward JD, Sullivan HG, Adams WB, Rosner MJ. Significance of intracranial hypertension in severe head injury. J Neurosurg 1977; 47: 503–516. 4. Miller JD, Butterworth JF, Gudeman SK, Faulkner JE, Choi SC, Selhorst JB, Harbison JW, Lutz HA, Young HF, Becker DP. Further experience with the management of severe head injury. J Neurosurg 1981; 54: 289–299. 5. Muizelaar JP, Marmarou A, Ward JD, Kontos HA, Choi SC, Becker DP, Gruemer H, Young HF. Adverse effects of prolonged hyperventilation in patients with severe head injury: randomized clinical trial. J Neurosurg 1991; 75: 731–739. 6. Cruz J. Combined continuous monitoring of systemic and cerebral oxygenation in acute brain injury: preliminary observations. Crit Care Med 1993; 21: 1225–32. 7. Cruz J, Miner M, Allen S, Alves W, Gennarelli T. Continuous monitoring of cerebral oxygenation in acute brain injury: injection of mannitol during hyperventilation. J Neurosurg 1990; 73: 725–730. 8. Cruz J, Miner M, Allen S, Alves W, Gennarelli T. Continuous monitoring of cerebral oxygenation in acute brain injury: assessment of cerebral hemodynamic reserve. Neurosurgery 1991; 29: 743–749. 9. Cruz J, Raps E, Hoffstad O, Jaggi J, Gennarelli T. Cerebral oxygenation monitoring. Crit Care Med 1983; 21: 1242–1246. 10. Robertson CS, Narayan RK, Gokaslan ZL, Pahwa R, Grossman RG, Caram P Jr, Allen E. Cerebral arteriovenous oxygen difference as an estimate of cerebral blood flow in comatose patients. J Neurosurg 1989; 70: 222–230. 11. Sheinberg M, Kanter M, Robertson CS, Contant C, Narayan R, Grossman RG. Continuous monitoring of venous oxygen saturation in head-injured patients. J Neurosurg 1992; 76: 212–217. 12. Gopinath SP, Robertson CS, Contant CF, Hayes C, Feldman Z, Narayan RK, Grossman RG. Jugular venous desaturation and outcome after head injury. J Neurol Neurosurg Psychiatry 1994; 57: 717–723. 13. Valadka AB, Gospinath SP, Contant CF, Uzura M, Robertson CS. Relationship of brain PO2 to outcome after severe head injury. Crit Care Med 1998; 26: 1576–1581. 14. Baker SP, O’Neill B, Haddon W, Long WP. The injury severity score: a method for describing patients with multiple injuries and evaluating emergency care. J Trauma 1974; 14: 187–9. 15. Baker SP, O’Neill B. The injury severity score: an update. J Trauma 1976; 16: 882–5. 16. Knaus W, Draper E, Wagner, D, Zimmermann J. APACHE II: A severity of disease classification system. Crit Care Med 1985; 13: 818–829. 17. Teasdale G, Jennett B. Assessment of coma and impaired consciousness. Lancet 1974; 2: 81–84. 18. Stocker R., Bernays R., Kossmann T, Imhof HG. Monitoring and treatment of acute head injury. In: Goris E, Trentz O (eds). The integrated approach to
© 2000 Harcourt Publishers Ltd
21.
22.
23. 24.
25. 26.
27.
28.
29. 30.
31.
32.
33. 34. 35.
36.
37.
trauma care. The first 24 hours. Berlin, Heidelberg and New York: Springer Verlag, 1995; 196–210. Stocker R, Fandino J, Kossmann T, Trentz O, Imhof HG. Experiences in primary intensive care. In: Diemath HE, Sommerauer J, von Wild KRH (eds). Brain protection in severe head injury. Munchen, Bern, Wien and New York: W. Zuckshwerdt Verlag, 1996; 60–66. Marshall L, Marshall S, Klauber MR, Berkum Clark M, Eisenberg H, Jane J, Luerssen T, Marmarou A, Foulkes M. A new classification of head injury based on computerized tomography. J Neurosurg 1991; 75: S14–20. Andrews PJ, Dearden NM, Miller JD. Jugular bulb cannulation technique and validation of a new continuous monitor. Br Journal of Anaesthesia 1991; 67: 553–558. Jakobsen M, Enevoldsen E. Retrograde catheterization of the right internal jugular vein for serial measurements of cerebral venous oxygen content. J of Cereb Blood Flow and Metab 1989; 9: 717–720. Jennett B, Bond M. Assessment of outcome after severe brain damage. Lancet 1975; 1: 480–484. Marmarou A, Anderson RL, Ward JD, Choi SC, Young HF. Impact of ICP instability and hypotension on outcome in patients with severe head injury. J Neurosurg 1991; 75: 559–566. Adams JH, Graham DI, Gennarelli TA. Diffuse axonal injury in head injury caused by a fall. Lancet 1984; 22: 1420–1422. Kalimo H, Rehcrona S, Soderfeldt B, Olsson Y, Siesjö BK. Brain lactic acidosis and ischemic cell damage: 2. Histopathology. J Cereb Blood Flow Metab 1981; 1: 313–327. Siesjo BK, Katsura K, Zhao Q, Folbergrova J, Pahlmark K, Siesjo P, Smith ML. Mechanism of secondary brain damage in global and focal ischemia: a speculative synthesis. J Neurotrauma 1995; 12: 943–956. Krieger D, Adams H, Schwarz S, Rieke K, Aschoff A, Hacke W. Prognostic and clinical relevance of pupillary responses, intracranial pressure monitoring, and brainstem auditory evoked potentials in comatose patients with acute supratentorial mass lesions. Crit Care Med 1993; 21: 1944–1950. Ropper AH. Lateral displacement of the brain and level of consciousness in patients with acute hemispheral mass. N Engl J Med 1986; 314: 953–958. Eisenberg HM, Gary HE Jr, Aldrich EF, Saydjari C, Turner B, Foulkes MA, Jane JA, Marmarou A, Marshall LF, Young HF. Initial CT findings in 753 patients with severe head injury. J Neurosurg 1990; 73: 688–698. Coplin WM, O’Keefe GE, Grady MS, Grant GA, March KS, Winn HR, Lam AM. Thrombotic, infectious, and procedural complications of the jugular bulb catheter in the intensive care unit. Neurosurgery 1997; 41: 101–109. Robertson CS, Grossman R, Goodman C, Narayan R. The predictive value of cerebral anaerobic metabolism with cerebral infarction after head injury. J Neurosurg 1987; 67: 361–368. Dolder E. Fortschritte in der Behandlung von schweren Schaedel-HirnTrauma-Verletzung? Unfallchirurgie 1983; 94: 116–121. Zangger P, Bossert W, Heinz C. Bedürfnisanalyse und Bilanz 5 Jahre nach schweren Hirnverletzungen. Z. Unfallchirur Vers Med 1991; 84: 91–94. Rabow L, DeSalles AF, Becker DP, Yang M, Kontos HA, Ward JD, Moulton RJ, Clifton G, Gruemer HD, Muizelaar JP. CSF brain creatine kinase levels and lactic acidosis in severe head injury. J Neurosurg 1986; 65: 625–629. Rehcrona S, Rosen I, Siesjö BK. Brain acidosis and ischemic cell damage: 1. Biochemistry and neurophysiology. J of Cer Blood Flow and Met 1981; 1: 297–311. Obrist W, Langfitt T, Jaggi J, Cruz J, Gennarelli T.: Cerebral blood flow and metabolism in comatose patients with acute head injury. J Neurosurg 1984; 61: 241–253.
Journal of Clinical Neuroscience (2000) 7(3), 226–233