- Email: [email protected]
Severe head injury among children: Prognostic factors and outcome
Injury, Int. J. Care Injured 40 (2009) 535–540
Contents lists available at ScienceDirect
Injury journal homepage: www.elsevier.com/locate/injury
Severe head injury among children: Prognostic factors and outcome Mabrouk Bahloul a,*, Chokri Ben Hamida a, Hedi Chelly a, Adel Chaari a, Hatem Kallel a, Hassen Dammak a, Noureddine Rekik a, Kamel Bahloul b, Kheireddine Ben Mahfoudh c, Mongia Hachicha d, Mounir Bouaziz a a
Department of Intensive Care, Habib Bourguiba University Hospital, Sfax, Tunisia Department of Neurosurgery, Habib Bourguiba University Hospital, Sfax, Tunisia Department of Radiology, Habib Bourguiba University Hospital, Sfax, Tunisia d Department of Paediatrics, Hedi Chaker University Hospital, Sfax, Tunisia b c
A R T I C L E I N F O
S U M M A R Y
Article history: Accepted 20 April 2008
Aim: To determine predictive factors of mortality among children after traumatic brain injury. Methods: A retrospective study over 8 years of 222 children with severe head injury (Glasgow Coma Scale score 8) admitted to a university hospital (Sfax, Tunisia). Basic demographic, clinical, biological and radiological data were recorded on admission and during intensive care unit stay. Results: The study included 163 boys (73.4%) and 59 girls, with mean age 7.54 3.8 years. The main cause of trauma was road traffic accident (75.7%). Mean Glasgow Coma Scale score was 6 1.5, mean Injury Severity Score (ISS) was 28.2 6.9, mean Paediatric Trauma Score (PTS) was 3.7 2.1 and mean Paediatric Risk of Mortality (PRISM) was 14.3 8.5; 54 children (24.3%) died. Univariate analysis showed that low PTS on admission, high ISS or PRISM, presence of shock or meningeal haemorrhage or bilateral mydriasis, and serum glucose > 10 mmol l 1 were associated with mortality rate. Multivariate analysis showed that factors associated with a poor prognosis were PRISM > 20 and bilateral mydriasis on admission. Conclusions: In Tunisia, head injury is a frequent cause of hospital admission and is most often due to road traffic accidents. Short-term prognosis is poor, with a high mortality rate (24.3%), and is influenced by demographic, clinical, radiological and biological factors. ß 2008 Elsevier Ltd. All rights reserved.
Keywords: Trauma Acute head injury Children Intensive care unit Motor vehicle accident
Introduction Traumatic brain injury is the most common cause of death and of acquired disability among children and young adults in developed countries; even when adequate treatment is provided, there is usually neuronal loss.10 The pathophysiology of this condition highlights the importance not only of the primary lesions, but also of secondary processes that may lead to cerebral hypoxia and ischaemia.44 Secondary brain damage is the leading cause of death in hospital after traumatic brain injury.26,44 Moreover, the outcome of childhood head trauma varies from centre to centre depending on the availability of modern neurosurgical and neuroradiological facilities and qualified expertise.38 In Tunisia, nearly 13,000 victims of motor vehicle accident are recorded annually and about 1500 of these die, according to the National Guard statistical data.1 Paediatric morbidity and mortal-
* Corresponding author at: Service de Re´animation Me´dicale, Hoˆpital Habib Bourguiba, Route el Ain Km 1, 3029 Sfax, Tunisia. Tel.: +216 98698267; fax: +216 74243427. E-mail address: [email protected] (M. Bahloul). 0020–1383/$ – see front matter ß 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.injury.2008.04.018
ity due to head trauma are increasing because of the high rate of road traffic accidents. Survivors are susceptible to irreversible neurological damage that represents an important socioeconomic problem.13,31 In the Sfax area (South Tunisia), everyone with severe traumatic head injury is admitted to our medicosurgical intensive care unit (ICU), where specific monitoring tools (jugular venous saturation, intracranial pressure monitoring and transcranial Doppler sonography) are, however, not available. The aim of the present study was to evaluate outcome of severe head injury among children referred to this unit, and to define simple predictive factors which could be used in routine practice in general ICUs as indicators of prognosis. Materials and methods This study was approved by an internal review board. Patients In this retrospective study, we included all consecutive patients with severe traumatic brain injury and Glasgow Coma Scale (GCS)
536
M. Bahloul et al. / Injury, Int. J. Care Injured 40 (2009) 535–540
score 8, aged less than 15 years and admitted to the ICU of Habib Bourguiba University Hospital during an 8-year period, from 1997 to 2004. The data were recorded from the clinical notes, with multiple contributors. Our department is a 22-bed medicosurgical ICU in a teaching hospital of 510 beds, which serves as a first-line medical centre for an urban population of one million inhabitants and as a referral centre for a larger population coming from southern Tunisia. The total number of admissions in our unit is about 1200 per year, and 3 beds are reserved for paediatric intensive care. The children in this series were admitted directly from the scene of the accident, within 6 h of injury. They were all examined and scored according to the GCS on arrival, and underwent computed cerebral tomography (CT) as soon as feasible.
stay, including the means and ranges of all daily Na+, potassium (K+) and blood sugar levels. In addition, we recorded the development of secondary systemic insults (SSIs) on admission and during ICU stay. SSIs were divided into subgroups: respiratory (hypoxaemia, hypercapnia, hypocapnia),37,39 circulatory (hypotension or arterial hypertension),9,40 metabolic/electrolytic (anaemia, hyper- or hypoglycaemia, hyponatraemia, diabetes insipidus)5,22,37 and hyperthermia. All complications during ICU stay were recorded: nosocomial infections,20 pneumonia,33 urinary tract infections,33 meningitis20 and septicaemia.4 Glasgow Outcome Score27 was estimated after hospital discharge by an ICU physician and a paediatric physician (H.C. and M.H. in most cases).
Methods
Categorical data were expressed in proportions, and subgroups (survival and death) were analysed by the chi-squared testing. Continuous variables were expressed as means (standard deviation) and subgroups were evaluated by Student’s t-test. In this analysis, to study the influence of age on outcome, we compared mean ages between survivors and non-survivors. In addition, we analysed the association of age 2 years and age < 5 years and outcome. Risk factors were evaluated by univariate analysis and multivariate analysis using a multiple logistic stepwise regression procedure. Odds ratios were estimated from the coefficients obtained, with 95% confidence intervals. PRISM score, PTS, ISS and GCS score were used to predict mortality and were analysed by means of receiver operating characteristic (ROC) curves. The area under the ROC curve estimated by the method of Hanley and McNeill23 provides a measure of overall mortality. For comparable data a p-value < 0.05 was considered significant.
The patients’ medical files were retrospectively reviewed, and the following data were collected: age, gender, vital signs (heart and respiratory rates, systolic and diastolic blood pressures), body temperature in 8C, GCS score,45 Injury Severity Score (ISS),2 Paediatric Trauma Score (PTS),46 Paediatric Risk of Mortality (PRISM) score,41 cause of injury, pupillary responses, motor deficit, presence of convulsions, use of mechanical ventilation, presence of shock or arterial hypotension,21 occurrence of cardiac arrest, fluid intake volume, brain CT result and use of catecholamines (dopamine, dobutamine and epinephrine). Before 2005 norepinephrine was not available in our ICU, so it was not used in our study. Biochemical parameters measured on admission and during the ICU stay included arterial blood gases and acid–base status, haemoglobin concentration, platelet count, serum glucose and sodium (Na+) levels, blood urea and urinary specific gravity. Plain radiographic studies of the neck were performed in all cases. Cranial CT was carried out in all but four cases (because of reduced availability of CT); for these four children brain magnetic resonance imaging (MRI) was performed on admission. The CT findings were axed according to the presence or absence of haematoma (whether extradural, subdural or intracerebral), meningeal haemorrhage, cerebral oedema, cerebral contusion, pneumocephalus, intracranial mass lesion and herniation. In addition, cranial CT results were stratified according to the Traumatic Coma Databank Computed Tomography Classification for Severe Head Injury.50 The classification was performed by a university radiologist (K.B.M.). Neurological status was assessed using the GCS score, at the site of accident and again on hospital arrival, before the use of sedatives but after resuscitation, i.e., the pre-intubation GCS used in our analysis. All the children underwent intubation and ventilation and received sedation with thiopental sodium 50 mg kg 1/day or fentanyl-midazolam as necessary. Those with diabetes mellitus or who received intravenous glucose-containing fluid intravenously were recorded as such. Corticosteroids were not administered for treatment of cerebral oedema. Following the protocol in our ICU, the bedhead was kept elevated in all cases and mannitol was used when raised intracranial pressure was suspected or CT showed cerebral oedema and/or herniation. Hypertonic saline is not used in our practice. Mild hyperventilation (PaCO2 = 30–35 mmHg) was maintained as in all cases of severe traumatic head injury requiring mechanical ventilation. In our practice, anticonvulsants are administered if seizures develop. Neither hypothermia therapy nor decompressive craniectomy were used our series. Therapies were directed by repeated CT. When extracranial pathology was suspected, appropriate investigations were performed. All clinical, biological and radiological parameters and relevant therapeutic measures were recorded on admission and during ICU
Statistical analysis
Results During the study period, 455 children were admitted to our ICU with traumatic head injury. Of these, 222 had GCS score 8 and were included in the study. This group represented 16.2% of all paediatric ICU admissions, 80.2% of paediatric post-traumatic cases and 2.3% of all ICU admissions. Transport and stabilisation of vital functions were performed by a pre-hospital team and/or firefighters in 43% of cases. However, in 57% of cases transport was undertaken by the child’s family. Of the whole group, 44.4% were from Sfax city and district and (57.6%) were referred from other hospitals in southern Tunisia. The study group included 163 boys (73.4%) and 59 girls (26.6), with a mean age of 7.54 3.8 years (range 0.3–15 years). Children aged less than 2 years, those between 3 and 5 years and those between 6 and 10 years represented 8.1, 25.6 and 38% of the total population, respectively. The demographic and clinical parameters on admission are shown in Table 1. Trauma was caused by road traffic accident in 75.7%, occurred at home in 23.9% or was due to assault in 0.4% of cases. In 51.8% extracranial pathology was present, including fracture of ribs or long bones (27.5%), and injuries to the face (18%), chest (9%), abdomen (16.7%), pelvis (3.6%) and spine (1.4%). Brain CT was performed on admission in 218 cases, and in 4 cases the brain was explored on admission with MRI because of non-availability of CT; 27 (12.2%) showed depressed skull fracture with brain contusions. All the children needed intubation, mechanical ventilation (mean duration 5 6.3 days) and sedation on admission, according to the protocols detailed above. On admission, 40 (18%) children needed craniotomy (evacuation of subdural haematoma for 5, evacuation of extradural haematoma for 13, lobectomy for 2, cerebrospinal fluid drainage
M. Bahloul et al. / Injury, Int. J. Care Injured 40 (2009) 535–540
537
Table 1 Demographic and clinical parameters of study population on admission
Table 3 Frequency of each secondary systemic insult
Parameter
Number (%)
Insult
Number (%)
– 163/59 – –
Hyperthermia Hyponatraemia Arterial hypotension Hyperglycaemia (11 mmol l 1) Hypoxaemia Hypercapnia (>45 mmHg) Hypocapnia (<28 mmHg) Hypoglycaemia (<2.8 mmol l 1) Anaemia (Hb < 8.5 g dl 1) Diabetes insipidus
125 69 63 48 25 23 17 3 42 4
Mean S.D.
Age (years) Gender M/F PTS ISS PRISM HR (beats/min) SBP (mmHg) Respiratory distress Shock Cardiac arrest Body temperature (8C) GCS score Anisocoria Bilateral mydriasis Motor deficit Convulsion Other injury Pathological antecedent
7.5 3.8 3.7 2.1 28.2 6.9 14.3 8.5 114 29 100 20 – – – 38.1 1.3 6 1.5 – – – – – –
– – 23 (10.4) 10 (4.5) 7(3.2) – – 44 (20) 32 (14.5) 32 (14.5) 34 (15.4) 115 (51.8) 19 (8.6)
Table 4 Outcome for children with severe head injury according to the Glasgow Outcome Scale
S.D., standard deviation; M/F, male/female; PTS, Paediatric Trauma Score; ISS, Injury Severity Score; PRISM, Paediatric Risk of Mortality score; HR, heart rhythm; SBP, systolic blood pressure; GCS, Glasgow Coma Scale.
for 2 and elevation of depressed skull fracture for 18). Extracranial surgery was required immediately in 9 cases, and in 13 was needed a few days later. The results of brain CT are presented in Table 2. According to Marshall tomographic grading, we recorded 15.3% type I, 48.1% type II, 13.1% type III, 3.2% type IV, 14.9% type V and 5.4% type VI severe head injuries. Normal cerebral CT was observed in 22 cases (10%) and, of this subgroup, 7 children had extracranial pathology. However, all 22 children had a GCS 8 requiring mechanical ventilation and ICU admission. During ICU stay, 170 children (77.2%) had complications: nosocomial infection in 43 (19.4%), pneumonia in 27 (12.2%), urinary tract infection in 14 (6.3%), meningitis in 4 (1.8%), septicaemia in 6 (2.7%) and inner ear infection or sinusitis in 6 (2.7%) cases. During ICU stay 63 children (28.4%) developed arterial hypotension requiring fluid resuscitation. Catecholamines were used for 15 children (6.7%): dopamine for 14 (6.3%), epinephrine for 3 (1.3%) and dobutamine for 2 (0. 9%). A total of 112 patients (50.5%) had rhabdomyolysis with serum creatine phosphokinase (CPK) >500 IU/l.19 Hyponatraemia (<130 mmol l 1) was present in 69 (31%), hypernatraemia (>145 mmol l 1) in 26 (11.7%), diabetes insipidus in 4 (1.8%), fat embolism in 2 (1%), stage III or IV pressure ulcer49 in 13 (5.8%) and gastrointestinal haemorrhage in 3 (1.4%) cases. During ICU stay, 197 children (88.7%) developed SSIs as shown in Table 3. Mean ICU stay was 9 15 days. A total of 54 children (24.3%) died; mortality was 51% in the first 24–48 h, 31.3% between 3 and 7
Table 2 Cerebral CT findings CT signs
Number (%)
Normal Meningeal haemorrhage Cerebral oedema Cerebral contusion Extradural haematoma Subdural haematoma Pneumocephalus Mass lesion Cerebral trunk injury Skull fracture Depressed skull fracture
22 102 85 88 17 39 22 8 12 96 27
(9.9) (45.9) (38.3) (39.6) (7.7) (17.6) (9.9) (3.6) (5.4) (43.2) (12.2)
(56.3) (31) (28.4) (21.6) (11.3) (10.4) (7.7) (1.4) (19) (1.8)
Outcome
Number (%)
Death Persistent vegetative state Severe disability Moderate disability Good recovery
54 4 8 39 117
(24.3) (1.8) (3.6) (17.6) (52.7)
days and 16.6% thereafter. Brain herniation (diagnosed clinically) was the main cause of death (90.7%); other causes of mortality included acute respiratory distress in 4% and nosocomial infection in 5.3%. Among the 168 survivors, 20 (11.9%) had functional motor deficit, 14 (8.3%) had subjective symptoms and 3 (1.8%) had post-traumatic seizures. The Glasgow Outcome Score (GOS) at a mean 8 months after hospital discharge (0.5–96 months) was: 54 deaths (24.3%), 4 vegetative states (1.8%) and 117 good recoveries (52.7%), as displayed in Table 4. Univariate analysis showed that low PTS on admission, high ISS, high PRISM score, presence of shock, meningeal haemorrhage, serum glucose >10 mmol l 1 and bilateral mydriasis were associated with mortality rate (Table 5). In our study, the mean age did not differ significantly between survivors and non-survivors Table 5 Factors associated with death in univariate analysis Factor
Survivors
Deaths
p-Value
Age (mean years S.D.) Age 2 years Age 5 years PTS (mean S.D.) ISS (mea S.D.) PRISM score (mean S.D.) GCS score Shock Bilateral mydriasis Anisocoria Neuro-vegetative disorders Meningeal haemorrhage Subdural haematoma Cerebral contusion Prothrombinaemia (%) Platelet count (106 ml 1) HCO3 (mmol l 1) Serum glucose >10 mmol l Serum K < 3.5 mmol l 1 Serum Na > 145 mmol l 1 Diabetes insipidus SSIs
8.4 3.9 8.3% 36% 42 27.6 6.7 11.4 5.7 6.5 1.4 4.8% 5.3% 18% 6.5% 39.3% 15.4% 38% 59 16 253 85 18.7 3.2 15.5% 38% 5.3% 0% 86.9%
7.3 3.7 7.4% 28% 2.9 2.2 30 7 23.6 9.3 5.2 1.6 18.5% 42.6% 24% 18.5% 66.7% 24% 44% 52 18 235 91 16.8 3 42.6% 61% 31.5% 7.4% 96.3%
0.070 0.820 0.280 0.001 0.020 <0.001 <0.001 <0.004 <0.001 0.480 0.009 <0.001 0.140 0.40 0.008 0.200 0.010 <0.0001 0.010 <0.0001 0.002 0.030
1
S.D., standard deviation; PTS, Paediatric Trauma Score; ISS, Injury Severity Score; PRISM, Paediatric Risk of Mortality; GCS, Glasgow Coma Scale; SSIs, secondary systemic insults.
538
M. Bahloul et al. / Injury, Int. J. Care Injured 40 (2009) 535–540
Figure 1. Mortality rate according to Glasgow Coma Scale score.
(p = 0.07). In addition, neither age 2 years nor age 5 years were associated with a poorer outcome. Finally, in our study a serum Na+ level >145 mmol l 1 on ICU admission (maximum level encountered was 154 mmol l 1) was significantly associated with a poorer outcome (p < 0.0001). Moreover, hypernatraemia was significantly associated with hyperglycaemia (p = 0.01), high value of PRISM score (23 10 vs. 13 7; p < 0.0001), high value of ISS (31.5 7 vs. 28 6; p = 0.009) and low value of GCS score on admission (5.3 1.8 vs. 6.2 1.4; p = 0.006). Multivariate analysis showed that factors associated with a poor prognosis were PRISM score > 20 (p < 0.0001; OR = 7.08) and presence of bilateral mydriasis on admission (p = 0.011; OR = 5.03). A significant association was found between PRISM score and mortality rate; this model had high discriminative power, with an area under the ROC curve at 0.87. In addition, as shown in Fig. 1, a low value of GCS score on admission was associated with a poor outcome. In fact, GCS score 6 was associated with death with a sensitivity of 56%, a specificity of 78% and an area under the ROC curve at 0.71. However, ISS and PTS did not discriminate sufficiently, with areas below the ROC curve at 0.57 and 0.63, respectively. Blood glucose level on admission was significantly higher among children who died (11.2 7 mmol l 1 vs. 8.7 4 mmol l 1; p = 0.001) and among those having an initial GCS score of 3 (12.7 10.9 mmol l 1 vs. 9 4 mmol l 1; p = 0.006) when compared with survivors and children with GCS score > 3 (Figs. 2 and 3).
Figure 3. Serum glucose on admission according to outcome. Black lines, medians; boxes, 25–75%; error bars, ranges.
Figure 4. Mortality rate according to Traumatic Coma Databank Classification of lesions.50
According to the Traumatic Coma Databank Classification, mortality rate was 8.8% in type I, 17.8% in type II, 44.8% in type III, 57% in type IV, 12.1% in type V and 83% in type VI cases (p < 0.001), as shown in Fig. 4. Moreover, as depicted in Fig. 5, mortality rate was narrowly related with number of developed SSIs (p < 0.0001), increasing from 17.4% among children with only one SSI to 66.7% among those having more than five SSIs. In all, 93 children received continuous infusion of thiopental sodium for the first few years after the trauma, and none after this.
Figure 2. Association between Glasgow Coma Scale score and serum glucose on admission.
Figure 5. Mortality rate according to number of secondary systemic insults developed in intensive care unit.
M. Bahloul et al. / Injury, Int. J. Care Injured 40 (2009) 535–540
The use of thiopental sodium was not significantly associated with a poor outcome (p = 0.64), nosocomial infection (p = 0.11) or consequence on blood pressure (p = 0.55). Discussion In the developed world, head injury is the most common cause of mortality and morbidity in childhood. However, several studies have reported that there is a fair chance of recovery after severe brain injury among children.12,29 In our study, extracranial pathology was present in 51.8% of cases and may explain in part the high frequency of rhabdomyolysis. Despite the unavailability of specific monitoring tools (jugular venous saturation, intracranial pressure monitoring or transcranial Doppler sonography) in our ICU, the mortality rate of 24.3% was somewhat similar to those reported in other studies,11,38,51 particularly in developed countries.11,48,51 Several investigators have stated that age is a good predictor of mortality in traumatic brain injury.31 Younger persons have a better chance of survival, tolerate longer periods of coma or decerebration better than older persons and have fewer lifethreatening complications. Mortality is higher at extreme ages of life.31,48 The influence of age on outcome among children with severe head injury is controversial.6,15,28,48 As in other studies,11,51 this factor was not significant in our series (p = 0.07). However, we cannot conclude that age has no effect on outcome; it is possible that, in this retrospective study, not all children received the same treatment. Our investigation was established during an 8-year period from 1997 to 2004 and, before January 2000, barbiturate therapy (thiopental sodium 50 mg kg 1/day) was used in all cases admitted for severe traumatic head injury requiring mechanical ventilation. This therapy was abandoned after that date. The GCS scoring system was described to quantify state of consciousness.45 It is a useful tool, but must be meticulously and carefully used. Several studies have shown that there is a good correlation between GCS score and neurological outcome.31,48,51 In our series, we found that GCS score was associated with mortality rate only in univariate analysis (p < 0.001). Recovery of pupillary reflexes is a useful predictor of general recovery after brain trauma.24,42 In a prospective study, Riter et al.42 found that, among patients with GCS score < 8, death or a vegetative state occurred in 39% of those with two reactive pupils, in 66% of those with one reactive pupil and in 85% of those with non-reactive pupils. Fearnside et al.17 showed a significant difference in mortality according to whether both pupils reacted or not (p < 0.05); in fact, pupil reactivity is related to cerebral blood flow.42 In our study, bilateral mydriasis was clearly associated (p = 0.01) with mortality rate in the multivariate analysis. In the ICU these indicators contribute to an objective evaluation of outcome. The ISS is a commonly used scoring system in traumatology.7 Some studies have not found ISS to be a good outcome predictor, even in cases with serious injuries;16 others have found it to be a good predictor of poor prognosis.24,36 We demonstrated a significant relationship between ISS and prognosis in univariate analysis. The PRISM score has been widely used as a severity score for critically ill children, in various clinical situations.41 The few studies that have focused on the PRISM score for child trauma victims7 have demonstrated that the score is an accurate tool for predicting outcome.6,7,47 In our study, we showed with multivariate analysis that the PRISM score was a reliable tool for predicting death, as demonstrated by the high value (0.87) of the area under ROC curve. Our results were in agreement with previous studies.7,8,41
539
CT results are assumed to reflect the seriousness of head injuries and predict clinical course. The influence of these types of cerebral damage on cranial pressure and mortality rate has been evaluated variously in the literature.34,35 However, some lesions appear to carry a poorer prognosis; in our study, the presence of meningeal haemorrhage was associated with high mortality only in univariate analysis. Eisenberg et al.14 found a good relationship between such damage, increased intracranial pressure and death in a study of 753 cases of severe head injury, and the presence of these lesions appeared to double mortality in two otherwise comparable groups. An explanation could be found in the presence of cerebral ischaemia, due to reduced cerebral blood flow in the acute stages and a slight increase to levels near the lower limit of normal in the subacute stages.18 In addition, according to the Traumatic Coma Databank Classification, the mortality rate is significantly higher among people with type VI lesions. This classification differentiates between cases with and without mass lesions and has become widely accepted for descriptive purposes; it is also increasingly used as a major predictor of outcome in traumatic brain injury of adults.25,32 However, to our knowledge our study is the first to establish a strong association between this classification and outcome among children. The negative influence of SSIs is well documented. In our study, the development of SSIs was associated with a poor prognosis in univariate analysis. In addition, the mortality rate was very much associated with the number of SSIs. A significant correlation between glucose levels, severity of head trauma, pupillary reaction and maximum intracranial pressure during the first 24 h after trauma has been widely reported.43 Our study showed that a high level of blood glucose on admission was associated with initial severity of trauma (GCS score) and poor outcome. This result was previously reported in other studies.3,30 Hyponatraemia is associated with cerebral oedema, which augments secondary injury as identified in our study. Moreover, therapeutic hypernatraemia in neurological trauma is considered neuroprotective (e.g., use of hypertonic saline). In our study, Na+ > 145 mmol l 1 on ICU admission (maximum level encountered was 154 mmol l 1) was related to poor outcome. This can be explained by the associated high frequency of hyperglycaemia and high risk among this group, characterised by a low GCS score and a high value of PRISM score and ISS on admission. Despite the initial severity of our cases (PRISM score 14.3 8.5, 100% of the children needing intubation with mechanical ventilation), and although specific monitoring tools (jugular venous saturation, intracranial pressure monitoring and transcranial Doppler sonography) are not available in our ICU, the mortality rate (24.3%) in our series seems to be within world standards. Our results suggest that overall mortality rate may not differ significantly with or without invasive monitoring. This study was retrospective, but we were able to define some simple variables that were predictive of a poor short-term prognosis and that were based on easily measurable clinical, CT, biochemical and laboratory parameters that may be used either at the scene of the accident or in the emergency department of any hospital with available facilities. According to the National Guard statistical data, in Tunisia seatbelts are used by only 22% of car drivers and only 60% of motorcyclists wear helmets.1 In addition, the number of motor vehicles on the roads has clearly increased in the last decade. These data may explain in part the high frequency of motor vehicle accidents as the main cause of severe head injury. It would appear that an opportunity does exist for campaigns to increase seatbelt and helmet use in our population and to respect the speed limits, thereby decreasing serious head injury. Finally, we must mention that our investigation has several limitations. First, our study included a series of children only, with
M. Bahloul et al. / Injury, Int. J. Care Injured 40 (2009) 535–540
540
severe injury only. Results cannot, therefore, be extrapolated to cases of mild and moderate injury or to adults. Second, all retrospective studies such as this suffer from incomplete or inconsistent information. Third, the retrospective nature of the case series made it difficult to perform extensive analysis. Fourth, in our ICU the diagnosis of raised intracranial pressure was made on clinical grounds (pupillary response, motor deficit, occurrence of convulsions) and brain CT results, and we think that inability to make direct measurements adversely affected our assessment. Finally, in our study the GOS was performed within a mean 8 months after hospital discharge. We must point out that recovery recorded as good might be followed by poor long-term functional outcome, as the further development of the child’s brain might deviate more and more from the normal developmental path. In summary, traumatic brain injury is the most common cause of death and acquired disability among children and young adults in developed countries. Motor vehicle accidents represent the main cause of such severe head injury, and short-term prognosis is poor; prognosis for a head-injured child who is still alive cannot be calculated at the roadside. Improved pre-hospital care, readily available multidisciplinary emergency teams, the establishment of regional trauma centres, and efforts to prevent and mitigate motor vehicle accidents should improve the prognosis of severe head injury among children. Conflict of interest None. References 1. Bahloul M, Chelly H, Ben Hmida M, et al. Prognosis of traumatic head injury in South Tunisia: a multivariate analysis of 437 cases. J Trauma 2004;57:255–61. 2. Baker SP, O’Neill B, Haddon Jr W, et al. The injury severity score: a method for describing patients with multiple injuries and evaluating emergency care. J Trauma 1974;14:187–96. 3. Bochicchio GV, Sung J, Joshi M, et al. Persistent hyperglycemia is predictive of outcome in critically ill trauma patients. J Trauma 2005;58:921–4. 4. Bone RC. Sepsis, the sepsis syndrome, multi-organ failure: a plea for comparable definitions. Ann Intern Med 1991;114:332–3. 5. Born JD, Hans P, Smitz S, et al. Syndrome of inappropriate secretion of antidiuretic hormone after severe head injury. Surg Neurol 1985;23:383–7. 6. Campbell CG, Kuehn SM, Richards PM, et al. Medical and cognitive outcome in children with traumatic brain injury. Can J Neurol Sci 2004;31:213–9. 7. Cantais E, Paut O, Giorgi R, et al. Evaluating the prognosis of multiple, severely traumatized children in the intensive care unit. Intensive Care Med 2001;27: 1511–7. 8. Castello FV, Cassano A, Gregory P, et al. The Pediatric Risk of Mortality (PRISM) Score and Injury Severity Score (ISS) for predicting resource utilization and outcome of intensive care in pediatric trauma. Crit Care Med 1999;27:985–8. 9. Chesnut RM, Marshall LF, Klauber MR, et al. The role of secondary brain injury in determining outcome from severe head injury. J Trauma 1993;34:216–22. 10. Chiaretti A, Piastra M, Polidori G, et al. Correlation between neurotrophic factor expression and outcome of children with severe traumatic brain injury. Intensive Care Med 2003;29:1329–38. 11. Chiaretti A, Piastra M, Pulitano S, et al. Prognostic factors and outcome of children with severe head injury: an 8-year experience. Childs Nerv Syst 2002;18:129–36. 12. Clifton GL, Kreutzer JS, Choi SC, et al. Relationship between Glasgow Outcome Scale and neuropsychological measures after brain injury. Neurosurgery 1993;33:34–9. 13. Diamond PT. Brain injury in the commonwealth of Virginia: an analysis of central registry data 1988–1993. Brain Int 1996;10:413–9. 14. Eisenberg HM, Gary Jr HE, Aldrich EF, et al. Initial CT findings in 753 patients with severe head injury. A report from the NIH Traumatic Coma Data Bank. J Neurosurg 1990;73:688–98. 15. Ewing-Cobbs L, Fletcher JM, Levin HS, et al. Longitudinal neuropsychological outcome in infants and preschoolers with traumatic brain injury. J Int Neuropsychol Soc 1997;3:581–91. 16. Fandino J, Stocker R, Prokop S, et al. Cerebral oxygenation and systemic trauma related factors determining neurological outcome after brain injury. J Clin Neurosci 2000;7:226–33. 17. Fearnside MR, Cook RJ, McDougall P, et al. The Westmead Head Injury Project outcome in severe head injury. A comparative analysis of pre-hospital, clinical and CT variables. Br J Neurosurg 1993;7:267–79.
18. Fukuda T, Mikoshiba M, Fukushima C, et al. Effects of traumatic subarachnoid haemorrhage on pathological properties in diffuse brain injury: a comparison with aneurysmal subarachnoid hemorrhage. No Shinkei Geka 1996;24: 723–31. 19. Gabow PA, Kaehny WD, Kelleher SP, et al. The spectrum of rhabdomyolysis. Medicine (Baltimore) 1982;61:141–52. 20. Garner JS, Jarvis WR, Emori TG, et al. CDC definitions for nosocomial infections. Am J Infect Control 1988;16:128–40. 21. Granry JC, Dube´ L, Terminassian A, Frebet E, Le Rolle T. Multimodal monitoring of head injuries in children. Ann Fr Anesth Reanim 2002;21:148–56. 22. Greger NG, Kirkland RT, Clayton GW, et al. Central diabetes insipidus 22 years’ experience. Am J Dis Child 1986;140:551–4. 23. Hanley JA, McNeil BJ. The meaning and use of the area under a receiver operating characteristic (ROC) curve. Radiology 1982;143:29–36. 24. Hill DA, Delaney LM, Roncal S, et al. A chi-square automatic interaction detection (CHAID) analysis of factors determining trauma outcomes. J Trauma 1997;42:62–6. 25. Hukkelhoven CW, Steyerberg EW, Habbema JD, et al. Predicting outcome after traumatic brain injury: development and validation of a prognostic score based on admission characteristics. J Neurotrauma 2005;22:1025–9. 26. Jamshid G. Traumatic brain injury (sminar). Lancet 2000;356:923–9. 27. Jennett B, Bond M. Assessment of outcome after severe brain damage. Lancet 1975;1:480–4. 28. Johnson DL, Krishnamurthy S. Severe pediatric head injury: myth, magic, and actual fact. Pediatr Neurosurg 1998;28:167–72. 29. Kumaraswamy N, Naziah A, Abdullah J, et al. Outcome of children with traumatic brain injury in rural Malaysia. J Clin Neurosci 2002;9:251–5. 30. Lam AM, Winn HR, Cullen BF, Sundling N. Hyperglycemia and neurological outcome in patients with head injury. J Neurosurg 1991;75:545–51. 31. Luerssen TG, Klauber MR, Marshall LF, et al. Outcome from head injury related to patient’s age. A longitudinal prospective study of adult and pediatric head injury. J Neurosurg 1988;68:409–16. 32. Maas AI, Hukkelhoven CW, Marshall LF, et al. Prediction of outcome in traumatic brain injury with computed tomographic characteristics: a comparison between the computed tomographic classification and combinations of computed tomographic predictors. Neurosurgery 2005;57:1173–82. 33. Marquette CH, Georges H, Wallet F, et al. Diagnostic efficiency of endotracheal aspirates with quantitative bacterial cultures in intubated patients with suspected pneumonia. Comparison with the protected specimen brush. Am Rev Respir Dis 1993;148:138–44. 34. Marshall LF, Marshall SB, Klauber MR, et al. The diagnosis of head injury requires a classification based on computed axial tomography. J Neurotrauma 1992;9:287–92. 35. Marshall LF, Toole BM, Bowers SA, et al. The National Traumatic Coma Data Bank. Part 2. Patients who talk and deteriorate: implications for treatment. J Neurosurg 1983;59:285–8. 36. Meldon SW, Reilly M, Drew BL, et al. Trauma in the very elderly: a communitybased study of outcomes at trauma and nontrauma centers. J Trauma 2002;52:79–84. 37. Miller JD, Sweet RC, Narayan R, Becker DP. Early insults to the injured brain. JAMA 1978;240:439–42. 38. Odebode TO, Abubakar AM. Childhood head injury: causes, outcome, and outcome predictors. A Nigerian perspective. Pediatr Surg Int 2004;20:348–52. 39. Pfenninger EG, Lindner KH. Arterial blood gases in patients with acute head injury at the accident site and upon hospital admission. Acta Anaesthesiol Scand 1991;35:148–52. 40. Pietropaoli JA, Rogers FB, Shackford SR, et al. The deleterious effects of intraoperative hypotension on outcome in patients with severe head injuries. J Trauma 1992;33:403–7. 41. Pollack MM, Patel KM, Ruttimann UE, et al. PRISM III: an updated pediatric risk of mortality score. Crit Care Med 1996;24:743–52. 42. Ritter AM, Muizelaar JP, Barnes T, et al. Brain stem blood flow, pupillary response, and outcome in patients with severe head injuries. Neurosurgery 1999;44:941–8. 43. Rovlias A, Kotsou S. The influence of hyperglycemia on neurological outcome in patients with severe head injury. Neurosurgery 2000;46:335–43. 44. Signorini DF, Andrews PJ, Jones PA, et al. Adding insult to injury: the prognostic value of early secondary insults for survival after traumatic brain injury. J Neurol Neurosur Psychiatry 1999;66:26–31. 45. Simpson D, Reilly P. Pediatric coma scale. Lancet 1982;2:450. 46. Tepas III JJ, Mollitt DL, Talbert JL, et al. The pediatric trauma score as a predictor of injury severity in the injured child. J Pediatr Surg 1987;22:14–8. 47. Thakker JC, Splaingard M, Zhu J, Babel K, Bresnahan J, Havens PL. Survival and functional outcome of children requiring endotracheal intubation during therapy for severe traumatic brain injury. Crit Care Med 1997;25:1396–401. 48. Tilford JM, Simpson PM, Yeh TS, et al. Variation in therapy and outcome for pediatric head trauma patients. Crit Care Med 2001;29:1056–61. 49. U.S. Department of Health and Human Services. AHCPR guidelines: treatment of pressure ulcers. AHCPR publication 95-0652. Washington, DC, USA: Department of Health and Human Services, 1994; p. 12–13. 50. Vos PE, van Voskuilen AC, Beems T, et al. Evaluation of the traumatic coma data bank computed tomography classification for severe head injury. J Neurotrauma 2001;18:649–55. 51. White JR, Farukhi Z, Bull C, et al. Predictors of outcome in severely head-injured children. Crit Care Med 2001;29:534–40.
Contents lists available at ScienceDirect
Injury journal homepage: www.elsevier.com/locate/injury
Severe head injury among children: Prognostic factors and outcome Mabrouk Bahloul a,*, Chokri Ben Hamida a, Hedi Chelly a, Adel Chaari a, Hatem Kallel a, Hassen Dammak a, Noureddine Rekik a, Kamel Bahloul b, Kheireddine Ben Mahfoudh c, Mongia Hachicha d, Mounir Bouaziz a a
Department of Intensive Care, Habib Bourguiba University Hospital, Sfax, Tunisia Department of Neurosurgery, Habib Bourguiba University Hospital, Sfax, Tunisia Department of Radiology, Habib Bourguiba University Hospital, Sfax, Tunisia d Department of Paediatrics, Hedi Chaker University Hospital, Sfax, Tunisia b c
A R T I C L E I N F O
S U M M A R Y
Article history: Accepted 20 April 2008
Aim: To determine predictive factors of mortality among children after traumatic brain injury. Methods: A retrospective study over 8 years of 222 children with severe head injury (Glasgow Coma Scale score 8) admitted to a university hospital (Sfax, Tunisia). Basic demographic, clinical, biological and radiological data were recorded on admission and during intensive care unit stay. Results: The study included 163 boys (73.4%) and 59 girls, with mean age 7.54 3.8 years. The main cause of trauma was road traffic accident (75.7%). Mean Glasgow Coma Scale score was 6 1.5, mean Injury Severity Score (ISS) was 28.2 6.9, mean Paediatric Trauma Score (PTS) was 3.7 2.1 and mean Paediatric Risk of Mortality (PRISM) was 14.3 8.5; 54 children (24.3%) died. Univariate analysis showed that low PTS on admission, high ISS or PRISM, presence of shock or meningeal haemorrhage or bilateral mydriasis, and serum glucose > 10 mmol l 1 were associated with mortality rate. Multivariate analysis showed that factors associated with a poor prognosis were PRISM > 20 and bilateral mydriasis on admission. Conclusions: In Tunisia, head injury is a frequent cause of hospital admission and is most often due to road traffic accidents. Short-term prognosis is poor, with a high mortality rate (24.3%), and is influenced by demographic, clinical, radiological and biological factors. ß 2008 Elsevier Ltd. All rights reserved.
Keywords: Trauma Acute head injury Children Intensive care unit Motor vehicle accident
Introduction Traumatic brain injury is the most common cause of death and of acquired disability among children and young adults in developed countries; even when adequate treatment is provided, there is usually neuronal loss.10 The pathophysiology of this condition highlights the importance not only of the primary lesions, but also of secondary processes that may lead to cerebral hypoxia and ischaemia.44 Secondary brain damage is the leading cause of death in hospital after traumatic brain injury.26,44 Moreover, the outcome of childhood head trauma varies from centre to centre depending on the availability of modern neurosurgical and neuroradiological facilities and qualified expertise.38 In Tunisia, nearly 13,000 victims of motor vehicle accident are recorded annually and about 1500 of these die, according to the National Guard statistical data.1 Paediatric morbidity and mortal-
* Corresponding author at: Service de Re´animation Me´dicale, Hoˆpital Habib Bourguiba, Route el Ain Km 1, 3029 Sfax, Tunisia. Tel.: +216 98698267; fax: +216 74243427. E-mail address: [email protected] (M. Bahloul). 0020–1383/$ – see front matter ß 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.injury.2008.04.018
ity due to head trauma are increasing because of the high rate of road traffic accidents. Survivors are susceptible to irreversible neurological damage that represents an important socioeconomic problem.13,31 In the Sfax area (South Tunisia), everyone with severe traumatic head injury is admitted to our medicosurgical intensive care unit (ICU), where specific monitoring tools (jugular venous saturation, intracranial pressure monitoring and transcranial Doppler sonography) are, however, not available. The aim of the present study was to evaluate outcome of severe head injury among children referred to this unit, and to define simple predictive factors which could be used in routine practice in general ICUs as indicators of prognosis. Materials and methods This study was approved by an internal review board. Patients In this retrospective study, we included all consecutive patients with severe traumatic brain injury and Glasgow Coma Scale (GCS)
536
M. Bahloul et al. / Injury, Int. J. Care Injured 40 (2009) 535–540
score 8, aged less than 15 years and admitted to the ICU of Habib Bourguiba University Hospital during an 8-year period, from 1997 to 2004. The data were recorded from the clinical notes, with multiple contributors. Our department is a 22-bed medicosurgical ICU in a teaching hospital of 510 beds, which serves as a first-line medical centre for an urban population of one million inhabitants and as a referral centre for a larger population coming from southern Tunisia. The total number of admissions in our unit is about 1200 per year, and 3 beds are reserved for paediatric intensive care. The children in this series were admitted directly from the scene of the accident, within 6 h of injury. They were all examined and scored according to the GCS on arrival, and underwent computed cerebral tomography (CT) as soon as feasible.
stay, including the means and ranges of all daily Na+, potassium (K+) and blood sugar levels. In addition, we recorded the development of secondary systemic insults (SSIs) on admission and during ICU stay. SSIs were divided into subgroups: respiratory (hypoxaemia, hypercapnia, hypocapnia),37,39 circulatory (hypotension or arterial hypertension),9,40 metabolic/electrolytic (anaemia, hyper- or hypoglycaemia, hyponatraemia, diabetes insipidus)5,22,37 and hyperthermia. All complications during ICU stay were recorded: nosocomial infections,20 pneumonia,33 urinary tract infections,33 meningitis20 and septicaemia.4 Glasgow Outcome Score27 was estimated after hospital discharge by an ICU physician and a paediatric physician (H.C. and M.H. in most cases).
Methods
Categorical data were expressed in proportions, and subgroups (survival and death) were analysed by the chi-squared testing. Continuous variables were expressed as means (standard deviation) and subgroups were evaluated by Student’s t-test. In this analysis, to study the influence of age on outcome, we compared mean ages between survivors and non-survivors. In addition, we analysed the association of age 2 years and age < 5 years and outcome. Risk factors were evaluated by univariate analysis and multivariate analysis using a multiple logistic stepwise regression procedure. Odds ratios were estimated from the coefficients obtained, with 95% confidence intervals. PRISM score, PTS, ISS and GCS score were used to predict mortality and were analysed by means of receiver operating characteristic (ROC) curves. The area under the ROC curve estimated by the method of Hanley and McNeill23 provides a measure of overall mortality. For comparable data a p-value < 0.05 was considered significant.
The patients’ medical files were retrospectively reviewed, and the following data were collected: age, gender, vital signs (heart and respiratory rates, systolic and diastolic blood pressures), body temperature in 8C, GCS score,45 Injury Severity Score (ISS),2 Paediatric Trauma Score (PTS),46 Paediatric Risk of Mortality (PRISM) score,41 cause of injury, pupillary responses, motor deficit, presence of convulsions, use of mechanical ventilation, presence of shock or arterial hypotension,21 occurrence of cardiac arrest, fluid intake volume, brain CT result and use of catecholamines (dopamine, dobutamine and epinephrine). Before 2005 norepinephrine was not available in our ICU, so it was not used in our study. Biochemical parameters measured on admission and during the ICU stay included arterial blood gases and acid–base status, haemoglobin concentration, platelet count, serum glucose and sodium (Na+) levels, blood urea and urinary specific gravity. Plain radiographic studies of the neck were performed in all cases. Cranial CT was carried out in all but four cases (because of reduced availability of CT); for these four children brain magnetic resonance imaging (MRI) was performed on admission. The CT findings were axed according to the presence or absence of haematoma (whether extradural, subdural or intracerebral), meningeal haemorrhage, cerebral oedema, cerebral contusion, pneumocephalus, intracranial mass lesion and herniation. In addition, cranial CT results were stratified according to the Traumatic Coma Databank Computed Tomography Classification for Severe Head Injury.50 The classification was performed by a university radiologist (K.B.M.). Neurological status was assessed using the GCS score, at the site of accident and again on hospital arrival, before the use of sedatives but after resuscitation, i.e., the pre-intubation GCS used in our analysis. All the children underwent intubation and ventilation and received sedation with thiopental sodium 50 mg kg 1/day or fentanyl-midazolam as necessary. Those with diabetes mellitus or who received intravenous glucose-containing fluid intravenously were recorded as such. Corticosteroids were not administered for treatment of cerebral oedema. Following the protocol in our ICU, the bedhead was kept elevated in all cases and mannitol was used when raised intracranial pressure was suspected or CT showed cerebral oedema and/or herniation. Hypertonic saline is not used in our practice. Mild hyperventilation (PaCO2 = 30–35 mmHg) was maintained as in all cases of severe traumatic head injury requiring mechanical ventilation. In our practice, anticonvulsants are administered if seizures develop. Neither hypothermia therapy nor decompressive craniectomy were used our series. Therapies were directed by repeated CT. When extracranial pathology was suspected, appropriate investigations were performed. All clinical, biological and radiological parameters and relevant therapeutic measures were recorded on admission and during ICU
Statistical analysis
Results During the study period, 455 children were admitted to our ICU with traumatic head injury. Of these, 222 had GCS score 8 and were included in the study. This group represented 16.2% of all paediatric ICU admissions, 80.2% of paediatric post-traumatic cases and 2.3% of all ICU admissions. Transport and stabilisation of vital functions were performed by a pre-hospital team and/or firefighters in 43% of cases. However, in 57% of cases transport was undertaken by the child’s family. Of the whole group, 44.4% were from Sfax city and district and (57.6%) were referred from other hospitals in southern Tunisia. The study group included 163 boys (73.4%) and 59 girls (26.6), with a mean age of 7.54 3.8 years (range 0.3–15 years). Children aged less than 2 years, those between 3 and 5 years and those between 6 and 10 years represented 8.1, 25.6 and 38% of the total population, respectively. The demographic and clinical parameters on admission are shown in Table 1. Trauma was caused by road traffic accident in 75.7%, occurred at home in 23.9% or was due to assault in 0.4% of cases. In 51.8% extracranial pathology was present, including fracture of ribs or long bones (27.5%), and injuries to the face (18%), chest (9%), abdomen (16.7%), pelvis (3.6%) and spine (1.4%). Brain CT was performed on admission in 218 cases, and in 4 cases the brain was explored on admission with MRI because of non-availability of CT; 27 (12.2%) showed depressed skull fracture with brain contusions. All the children needed intubation, mechanical ventilation (mean duration 5 6.3 days) and sedation on admission, according to the protocols detailed above. On admission, 40 (18%) children needed craniotomy (evacuation of subdural haematoma for 5, evacuation of extradural haematoma for 13, lobectomy for 2, cerebrospinal fluid drainage
M. Bahloul et al. / Injury, Int. J. Care Injured 40 (2009) 535–540
537
Table 1 Demographic and clinical parameters of study population on admission
Table 3 Frequency of each secondary systemic insult
Parameter
Number (%)
Insult
Number (%)
– 163/59 – –
Hyperthermia Hyponatraemia Arterial hypotension Hyperglycaemia (11 mmol l 1) Hypoxaemia Hypercapnia (>45 mmHg) Hypocapnia (<28 mmHg) Hypoglycaemia (<2.8 mmol l 1) Anaemia (Hb < 8.5 g dl 1) Diabetes insipidus
125 69 63 48 25 23 17 3 42 4
Mean S.D.
Age (years) Gender M/F PTS ISS PRISM HR (beats/min) SBP (mmHg) Respiratory distress Shock Cardiac arrest Body temperature (8C) GCS score Anisocoria Bilateral mydriasis Motor deficit Convulsion Other injury Pathological antecedent
7.5 3.8 3.7 2.1 28.2 6.9 14.3 8.5 114 29 100 20 – – – 38.1 1.3 6 1.5 – – – – – –
– – 23 (10.4) 10 (4.5) 7(3.2) – – 44 (20) 32 (14.5) 32 (14.5) 34 (15.4) 115 (51.8) 19 (8.6)
Table 4 Outcome for children with severe head injury according to the Glasgow Outcome Scale
S.D., standard deviation; M/F, male/female; PTS, Paediatric Trauma Score; ISS, Injury Severity Score; PRISM, Paediatric Risk of Mortality score; HR, heart rhythm; SBP, systolic blood pressure; GCS, Glasgow Coma Scale.
for 2 and elevation of depressed skull fracture for 18). Extracranial surgery was required immediately in 9 cases, and in 13 was needed a few days later. The results of brain CT are presented in Table 2. According to Marshall tomographic grading, we recorded 15.3% type I, 48.1% type II, 13.1% type III, 3.2% type IV, 14.9% type V and 5.4% type VI severe head injuries. Normal cerebral CT was observed in 22 cases (10%) and, of this subgroup, 7 children had extracranial pathology. However, all 22 children had a GCS 8 requiring mechanical ventilation and ICU admission. During ICU stay, 170 children (77.2%) had complications: nosocomial infection in 43 (19.4%), pneumonia in 27 (12.2%), urinary tract infection in 14 (6.3%), meningitis in 4 (1.8%), septicaemia in 6 (2.7%) and inner ear infection or sinusitis in 6 (2.7%) cases. During ICU stay 63 children (28.4%) developed arterial hypotension requiring fluid resuscitation. Catecholamines were used for 15 children (6.7%): dopamine for 14 (6.3%), epinephrine for 3 (1.3%) and dobutamine for 2 (0. 9%). A total of 112 patients (50.5%) had rhabdomyolysis with serum creatine phosphokinase (CPK) >500 IU/l.19 Hyponatraemia (<130 mmol l 1) was present in 69 (31%), hypernatraemia (>145 mmol l 1) in 26 (11.7%), diabetes insipidus in 4 (1.8%), fat embolism in 2 (1%), stage III or IV pressure ulcer49 in 13 (5.8%) and gastrointestinal haemorrhage in 3 (1.4%) cases. During ICU stay, 197 children (88.7%) developed SSIs as shown in Table 3. Mean ICU stay was 9 15 days. A total of 54 children (24.3%) died; mortality was 51% in the first 24–48 h, 31.3% between 3 and 7
Table 2 Cerebral CT findings CT signs
Number (%)
Normal Meningeal haemorrhage Cerebral oedema Cerebral contusion Extradural haematoma Subdural haematoma Pneumocephalus Mass lesion Cerebral trunk injury Skull fracture Depressed skull fracture
22 102 85 88 17 39 22 8 12 96 27
(9.9) (45.9) (38.3) (39.6) (7.7) (17.6) (9.9) (3.6) (5.4) (43.2) (12.2)
(56.3) (31) (28.4) (21.6) (11.3) (10.4) (7.7) (1.4) (19) (1.8)
Outcome
Number (%)
Death Persistent vegetative state Severe disability Moderate disability Good recovery
54 4 8 39 117
(24.3) (1.8) (3.6) (17.6) (52.7)
days and 16.6% thereafter. Brain herniation (diagnosed clinically) was the main cause of death (90.7%); other causes of mortality included acute respiratory distress in 4% and nosocomial infection in 5.3%. Among the 168 survivors, 20 (11.9%) had functional motor deficit, 14 (8.3%) had subjective symptoms and 3 (1.8%) had post-traumatic seizures. The Glasgow Outcome Score (GOS) at a mean 8 months after hospital discharge (0.5–96 months) was: 54 deaths (24.3%), 4 vegetative states (1.8%) and 117 good recoveries (52.7%), as displayed in Table 4. Univariate analysis showed that low PTS on admission, high ISS, high PRISM score, presence of shock, meningeal haemorrhage, serum glucose >10 mmol l 1 and bilateral mydriasis were associated with mortality rate (Table 5). In our study, the mean age did not differ significantly between survivors and non-survivors Table 5 Factors associated with death in univariate analysis Factor
Survivors
Deaths
p-Value
Age (mean years S.D.) Age 2 years Age 5 years PTS (mean S.D.) ISS (mea S.D.) PRISM score (mean S.D.) GCS score Shock Bilateral mydriasis Anisocoria Neuro-vegetative disorders Meningeal haemorrhage Subdural haematoma Cerebral contusion Prothrombinaemia (%) Platelet count (106 ml 1) HCO3 (mmol l 1) Serum glucose >10 mmol l Serum K < 3.5 mmol l 1 Serum Na > 145 mmol l 1 Diabetes insipidus SSIs
8.4 3.9 8.3% 36% 42 27.6 6.7 11.4 5.7 6.5 1.4 4.8% 5.3% 18% 6.5% 39.3% 15.4% 38% 59 16 253 85 18.7 3.2 15.5% 38% 5.3% 0% 86.9%
7.3 3.7 7.4% 28% 2.9 2.2 30 7 23.6 9.3 5.2 1.6 18.5% 42.6% 24% 18.5% 66.7% 24% 44% 52 18 235 91 16.8 3 42.6% 61% 31.5% 7.4% 96.3%
0.070 0.820 0.280 0.001 0.020 <0.001 <0.001 <0.004 <0.001 0.480 0.009 <0.001 0.140 0.40 0.008 0.200 0.010 <0.0001 0.010 <0.0001 0.002 0.030
1
S.D., standard deviation; PTS, Paediatric Trauma Score; ISS, Injury Severity Score; PRISM, Paediatric Risk of Mortality; GCS, Glasgow Coma Scale; SSIs, secondary systemic insults.
538
M. Bahloul et al. / Injury, Int. J. Care Injured 40 (2009) 535–540
Figure 1. Mortality rate according to Glasgow Coma Scale score.
(p = 0.07). In addition, neither age 2 years nor age 5 years were associated with a poorer outcome. Finally, in our study a serum Na+ level >145 mmol l 1 on ICU admission (maximum level encountered was 154 mmol l 1) was significantly associated with a poorer outcome (p < 0.0001). Moreover, hypernatraemia was significantly associated with hyperglycaemia (p = 0.01), high value of PRISM score (23 10 vs. 13 7; p < 0.0001), high value of ISS (31.5 7 vs. 28 6; p = 0.009) and low value of GCS score on admission (5.3 1.8 vs. 6.2 1.4; p = 0.006). Multivariate analysis showed that factors associated with a poor prognosis were PRISM score > 20 (p < 0.0001; OR = 7.08) and presence of bilateral mydriasis on admission (p = 0.011; OR = 5.03). A significant association was found between PRISM score and mortality rate; this model had high discriminative power, with an area under the ROC curve at 0.87. In addition, as shown in Fig. 1, a low value of GCS score on admission was associated with a poor outcome. In fact, GCS score 6 was associated with death with a sensitivity of 56%, a specificity of 78% and an area under the ROC curve at 0.71. However, ISS and PTS did not discriminate sufficiently, with areas below the ROC curve at 0.57 and 0.63, respectively. Blood glucose level on admission was significantly higher among children who died (11.2 7 mmol l 1 vs. 8.7 4 mmol l 1; p = 0.001) and among those having an initial GCS score of 3 (12.7 10.9 mmol l 1 vs. 9 4 mmol l 1; p = 0.006) when compared with survivors and children with GCS score > 3 (Figs. 2 and 3).
Figure 3. Serum glucose on admission according to outcome. Black lines, medians; boxes, 25–75%; error bars, ranges.
Figure 4. Mortality rate according to Traumatic Coma Databank Classification of lesions.50
According to the Traumatic Coma Databank Classification, mortality rate was 8.8% in type I, 17.8% in type II, 44.8% in type III, 57% in type IV, 12.1% in type V and 83% in type VI cases (p < 0.001), as shown in Fig. 4. Moreover, as depicted in Fig. 5, mortality rate was narrowly related with number of developed SSIs (p < 0.0001), increasing from 17.4% among children with only one SSI to 66.7% among those having more than five SSIs. In all, 93 children received continuous infusion of thiopental sodium for the first few years after the trauma, and none after this.
Figure 2. Association between Glasgow Coma Scale score and serum glucose on admission.
Figure 5. Mortality rate according to number of secondary systemic insults developed in intensive care unit.
M. Bahloul et al. / Injury, Int. J. Care Injured 40 (2009) 535–540
The use of thiopental sodium was not significantly associated with a poor outcome (p = 0.64), nosocomial infection (p = 0.11) or consequence on blood pressure (p = 0.55). Discussion In the developed world, head injury is the most common cause of mortality and morbidity in childhood. However, several studies have reported that there is a fair chance of recovery after severe brain injury among children.12,29 In our study, extracranial pathology was present in 51.8% of cases and may explain in part the high frequency of rhabdomyolysis. Despite the unavailability of specific monitoring tools (jugular venous saturation, intracranial pressure monitoring or transcranial Doppler sonography) in our ICU, the mortality rate of 24.3% was somewhat similar to those reported in other studies,11,38,51 particularly in developed countries.11,48,51 Several investigators have stated that age is a good predictor of mortality in traumatic brain injury.31 Younger persons have a better chance of survival, tolerate longer periods of coma or decerebration better than older persons and have fewer lifethreatening complications. Mortality is higher at extreme ages of life.31,48 The influence of age on outcome among children with severe head injury is controversial.6,15,28,48 As in other studies,11,51 this factor was not significant in our series (p = 0.07). However, we cannot conclude that age has no effect on outcome; it is possible that, in this retrospective study, not all children received the same treatment. Our investigation was established during an 8-year period from 1997 to 2004 and, before January 2000, barbiturate therapy (thiopental sodium 50 mg kg 1/day) was used in all cases admitted for severe traumatic head injury requiring mechanical ventilation. This therapy was abandoned after that date. The GCS scoring system was described to quantify state of consciousness.45 It is a useful tool, but must be meticulously and carefully used. Several studies have shown that there is a good correlation between GCS score and neurological outcome.31,48,51 In our series, we found that GCS score was associated with mortality rate only in univariate analysis (p < 0.001). Recovery of pupillary reflexes is a useful predictor of general recovery after brain trauma.24,42 In a prospective study, Riter et al.42 found that, among patients with GCS score < 8, death or a vegetative state occurred in 39% of those with two reactive pupils, in 66% of those with one reactive pupil and in 85% of those with non-reactive pupils. Fearnside et al.17 showed a significant difference in mortality according to whether both pupils reacted or not (p < 0.05); in fact, pupil reactivity is related to cerebral blood flow.42 In our study, bilateral mydriasis was clearly associated (p = 0.01) with mortality rate in the multivariate analysis. In the ICU these indicators contribute to an objective evaluation of outcome. The ISS is a commonly used scoring system in traumatology.7 Some studies have not found ISS to be a good outcome predictor, even in cases with serious injuries;16 others have found it to be a good predictor of poor prognosis.24,36 We demonstrated a significant relationship between ISS and prognosis in univariate analysis. The PRISM score has been widely used as a severity score for critically ill children, in various clinical situations.41 The few studies that have focused on the PRISM score for child trauma victims7 have demonstrated that the score is an accurate tool for predicting outcome.6,7,47 In our study, we showed with multivariate analysis that the PRISM score was a reliable tool for predicting death, as demonstrated by the high value (0.87) of the area under ROC curve. Our results were in agreement with previous studies.7,8,41
539
CT results are assumed to reflect the seriousness of head injuries and predict clinical course. The influence of these types of cerebral damage on cranial pressure and mortality rate has been evaluated variously in the literature.34,35 However, some lesions appear to carry a poorer prognosis; in our study, the presence of meningeal haemorrhage was associated with high mortality only in univariate analysis. Eisenberg et al.14 found a good relationship between such damage, increased intracranial pressure and death in a study of 753 cases of severe head injury, and the presence of these lesions appeared to double mortality in two otherwise comparable groups. An explanation could be found in the presence of cerebral ischaemia, due to reduced cerebral blood flow in the acute stages and a slight increase to levels near the lower limit of normal in the subacute stages.18 In addition, according to the Traumatic Coma Databank Classification, the mortality rate is significantly higher among people with type VI lesions. This classification differentiates between cases with and without mass lesions and has become widely accepted for descriptive purposes; it is also increasingly used as a major predictor of outcome in traumatic brain injury of adults.25,32 However, to our knowledge our study is the first to establish a strong association between this classification and outcome among children. The negative influence of SSIs is well documented. In our study, the development of SSIs was associated with a poor prognosis in univariate analysis. In addition, the mortality rate was very much associated with the number of SSIs. A significant correlation between glucose levels, severity of head trauma, pupillary reaction and maximum intracranial pressure during the first 24 h after trauma has been widely reported.43 Our study showed that a high level of blood glucose on admission was associated with initial severity of trauma (GCS score) and poor outcome. This result was previously reported in other studies.3,30 Hyponatraemia is associated with cerebral oedema, which augments secondary injury as identified in our study. Moreover, therapeutic hypernatraemia in neurological trauma is considered neuroprotective (e.g., use of hypertonic saline). In our study, Na+ > 145 mmol l 1 on ICU admission (maximum level encountered was 154 mmol l 1) was related to poor outcome. This can be explained by the associated high frequency of hyperglycaemia and high risk among this group, characterised by a low GCS score and a high value of PRISM score and ISS on admission. Despite the initial severity of our cases (PRISM score 14.3 8.5, 100% of the children needing intubation with mechanical ventilation), and although specific monitoring tools (jugular venous saturation, intracranial pressure monitoring and transcranial Doppler sonography) are not available in our ICU, the mortality rate (24.3%) in our series seems to be within world standards. Our results suggest that overall mortality rate may not differ significantly with or without invasive monitoring. This study was retrospective, but we were able to define some simple variables that were predictive of a poor short-term prognosis and that were based on easily measurable clinical, CT, biochemical and laboratory parameters that may be used either at the scene of the accident or in the emergency department of any hospital with available facilities. According to the National Guard statistical data, in Tunisia seatbelts are used by only 22% of car drivers and only 60% of motorcyclists wear helmets.1 In addition, the number of motor vehicles on the roads has clearly increased in the last decade. These data may explain in part the high frequency of motor vehicle accidents as the main cause of severe head injury. It would appear that an opportunity does exist for campaigns to increase seatbelt and helmet use in our population and to respect the speed limits, thereby decreasing serious head injury. Finally, we must mention that our investigation has several limitations. First, our study included a series of children only, with
M. Bahloul et al. / Injury, Int. J. Care Injured 40 (2009) 535–540
540
severe injury only. Results cannot, therefore, be extrapolated to cases of mild and moderate injury or to adults. Second, all retrospective studies such as this suffer from incomplete or inconsistent information. Third, the retrospective nature of the case series made it difficult to perform extensive analysis. Fourth, in our ICU the diagnosis of raised intracranial pressure was made on clinical grounds (pupillary response, motor deficit, occurrence of convulsions) and brain CT results, and we think that inability to make direct measurements adversely affected our assessment. Finally, in our study the GOS was performed within a mean 8 months after hospital discharge. We must point out that recovery recorded as good might be followed by poor long-term functional outcome, as the further development of the child’s brain might deviate more and more from the normal developmental path. In summary, traumatic brain injury is the most common cause of death and acquired disability among children and young adults in developed countries. Motor vehicle accidents represent the main cause of such severe head injury, and short-term prognosis is poor; prognosis for a head-injured child who is still alive cannot be calculated at the roadside. Improved pre-hospital care, readily available multidisciplinary emergency teams, the establishment of regional trauma centres, and efforts to prevent and mitigate motor vehicle accidents should improve the prognosis of severe head injury among children. Conflict of interest None. References 1. Bahloul M, Chelly H, Ben Hmida M, et al. Prognosis of traumatic head injury in South Tunisia: a multivariate analysis of 437 cases. J Trauma 2004;57:255–61. 2. Baker SP, O’Neill B, Haddon Jr W, et al. The injury severity score: a method for describing patients with multiple injuries and evaluating emergency care. J Trauma 1974;14:187–96. 3. Bochicchio GV, Sung J, Joshi M, et al. Persistent hyperglycemia is predictive of outcome in critically ill trauma patients. J Trauma 2005;58:921–4. 4. Bone RC. Sepsis, the sepsis syndrome, multi-organ failure: a plea for comparable definitions. Ann Intern Med 1991;114:332–3. 5. Born JD, Hans P, Smitz S, et al. Syndrome of inappropriate secretion of antidiuretic hormone after severe head injury. Surg Neurol 1985;23:383–7. 6. Campbell CG, Kuehn SM, Richards PM, et al. Medical and cognitive outcome in children with traumatic brain injury. Can J Neurol Sci 2004;31:213–9. 7. Cantais E, Paut O, Giorgi R, et al. Evaluating the prognosis of multiple, severely traumatized children in the intensive care unit. Intensive Care Med 2001;27: 1511–7. 8. Castello FV, Cassano A, Gregory P, et al. The Pediatric Risk of Mortality (PRISM) Score and Injury Severity Score (ISS) for predicting resource utilization and outcome of intensive care in pediatric trauma. Crit Care Med 1999;27:985–8. 9. Chesnut RM, Marshall LF, Klauber MR, et al. The role of secondary brain injury in determining outcome from severe head injury. J Trauma 1993;34:216–22. 10. Chiaretti A, Piastra M, Polidori G, et al. Correlation between neurotrophic factor expression and outcome of children with severe traumatic brain injury. Intensive Care Med 2003;29:1329–38. 11. Chiaretti A, Piastra M, Pulitano S, et al. Prognostic factors and outcome of children with severe head injury: an 8-year experience. Childs Nerv Syst 2002;18:129–36. 12. Clifton GL, Kreutzer JS, Choi SC, et al. Relationship between Glasgow Outcome Scale and neuropsychological measures after brain injury. Neurosurgery 1993;33:34–9. 13. Diamond PT. Brain injury in the commonwealth of Virginia: an analysis of central registry data 1988–1993. Brain Int 1996;10:413–9. 14. Eisenberg HM, Gary Jr HE, Aldrich EF, et al. Initial CT findings in 753 patients with severe head injury. A report from the NIH Traumatic Coma Data Bank. J Neurosurg 1990;73:688–98. 15. Ewing-Cobbs L, Fletcher JM, Levin HS, et al. Longitudinal neuropsychological outcome in infants and preschoolers with traumatic brain injury. J Int Neuropsychol Soc 1997;3:581–91. 16. Fandino J, Stocker R, Prokop S, et al. Cerebral oxygenation and systemic trauma related factors determining neurological outcome after brain injury. J Clin Neurosci 2000;7:226–33. 17. Fearnside MR, Cook RJ, McDougall P, et al. The Westmead Head Injury Project outcome in severe head injury. A comparative analysis of pre-hospital, clinical and CT variables. Br J Neurosurg 1993;7:267–79.
18. Fukuda T, Mikoshiba M, Fukushima C, et al. Effects of traumatic subarachnoid haemorrhage on pathological properties in diffuse brain injury: a comparison with aneurysmal subarachnoid hemorrhage. No Shinkei Geka 1996;24: 723–31. 19. Gabow PA, Kaehny WD, Kelleher SP, et al. The spectrum of rhabdomyolysis. Medicine (Baltimore) 1982;61:141–52. 20. Garner JS, Jarvis WR, Emori TG, et al. CDC definitions for nosocomial infections. Am J Infect Control 1988;16:128–40. 21. Granry JC, Dube´ L, Terminassian A, Frebet E, Le Rolle T. Multimodal monitoring of head injuries in children. Ann Fr Anesth Reanim 2002;21:148–56. 22. Greger NG, Kirkland RT, Clayton GW, et al. Central diabetes insipidus 22 years’ experience. Am J Dis Child 1986;140:551–4. 23. Hanley JA, McNeil BJ. The meaning and use of the area under a receiver operating characteristic (ROC) curve. Radiology 1982;143:29–36. 24. Hill DA, Delaney LM, Roncal S, et al. A chi-square automatic interaction detection (CHAID) analysis of factors determining trauma outcomes. J Trauma 1997;42:62–6. 25. Hukkelhoven CW, Steyerberg EW, Habbema JD, et al. Predicting outcome after traumatic brain injury: development and validation of a prognostic score based on admission characteristics. J Neurotrauma 2005;22:1025–9. 26. Jamshid G. Traumatic brain injury (sminar). Lancet 2000;356:923–9. 27. Jennett B, Bond M. Assessment of outcome after severe brain damage. Lancet 1975;1:480–4. 28. Johnson DL, Krishnamurthy S. Severe pediatric head injury: myth, magic, and actual fact. Pediatr Neurosurg 1998;28:167–72. 29. Kumaraswamy N, Naziah A, Abdullah J, et al. Outcome of children with traumatic brain injury in rural Malaysia. J Clin Neurosci 2002;9:251–5. 30. Lam AM, Winn HR, Cullen BF, Sundling N. Hyperglycemia and neurological outcome in patients with head injury. J Neurosurg 1991;75:545–51. 31. Luerssen TG, Klauber MR, Marshall LF, et al. Outcome from head injury related to patient’s age. A longitudinal prospective study of adult and pediatric head injury. J Neurosurg 1988;68:409–16. 32. Maas AI, Hukkelhoven CW, Marshall LF, et al. Prediction of outcome in traumatic brain injury with computed tomographic characteristics: a comparison between the computed tomographic classification and combinations of computed tomographic predictors. Neurosurgery 2005;57:1173–82. 33. Marquette CH, Georges H, Wallet F, et al. Diagnostic efficiency of endotracheal aspirates with quantitative bacterial cultures in intubated patients with suspected pneumonia. Comparison with the protected specimen brush. Am Rev Respir Dis 1993;148:138–44. 34. Marshall LF, Marshall SB, Klauber MR, et al. The diagnosis of head injury requires a classification based on computed axial tomography. J Neurotrauma 1992;9:287–92. 35. Marshall LF, Toole BM, Bowers SA, et al. The National Traumatic Coma Data Bank. Part 2. Patients who talk and deteriorate: implications for treatment. J Neurosurg 1983;59:285–8. 36. Meldon SW, Reilly M, Drew BL, et al. Trauma in the very elderly: a communitybased study of outcomes at trauma and nontrauma centers. J Trauma 2002;52:79–84. 37. Miller JD, Sweet RC, Narayan R, Becker DP. Early insults to the injured brain. JAMA 1978;240:439–42. 38. Odebode TO, Abubakar AM. Childhood head injury: causes, outcome, and outcome predictors. A Nigerian perspective. Pediatr Surg Int 2004;20:348–52. 39. Pfenninger EG, Lindner KH. Arterial blood gases in patients with acute head injury at the accident site and upon hospital admission. Acta Anaesthesiol Scand 1991;35:148–52. 40. Pietropaoli JA, Rogers FB, Shackford SR, et al. The deleterious effects of intraoperative hypotension on outcome in patients with severe head injuries. J Trauma 1992;33:403–7. 41. Pollack MM, Patel KM, Ruttimann UE, et al. PRISM III: an updated pediatric risk of mortality score. Crit Care Med 1996;24:743–52. 42. Ritter AM, Muizelaar JP, Barnes T, et al. Brain stem blood flow, pupillary response, and outcome in patients with severe head injuries. Neurosurgery 1999;44:941–8. 43. Rovlias A, Kotsou S. The influence of hyperglycemia on neurological outcome in patients with severe head injury. Neurosurgery 2000;46:335–43. 44. Signorini DF, Andrews PJ, Jones PA, et al. Adding insult to injury: the prognostic value of early secondary insults for survival after traumatic brain injury. J Neurol Neurosur Psychiatry 1999;66:26–31. 45. Simpson D, Reilly P. Pediatric coma scale. Lancet 1982;2:450. 46. Tepas III JJ, Mollitt DL, Talbert JL, et al. The pediatric trauma score as a predictor of injury severity in the injured child. J Pediatr Surg 1987;22:14–8. 47. Thakker JC, Splaingard M, Zhu J, Babel K, Bresnahan J, Havens PL. Survival and functional outcome of children requiring endotracheal intubation during therapy for severe traumatic brain injury. Crit Care Med 1997;25:1396–401. 48. Tilford JM, Simpson PM, Yeh TS, et al. Variation in therapy and outcome for pediatric head trauma patients. Crit Care Med 2001;29:1056–61. 49. U.S. Department of Health and Human Services. AHCPR guidelines: treatment of pressure ulcers. AHCPR publication 95-0652. Washington, DC, USA: Department of Health and Human Services, 1994; p. 12–13. 50. Vos PE, van Voskuilen AC, Beems T, et al. Evaluation of the traumatic coma data bank computed tomography classification for severe head injury. J Neurotrauma 2001;18:649–55. 51. White JR, Farukhi Z, Bull C, et al. Predictors of outcome in severely head-injured children. Crit Care Med 2001;29:534–40.