One hundred and thirty-five patients were enrolled in the study. One hundred and fourteen (84.44%) were males and 21 (15.55%) were females (SR, 5.42). Their mean age was 32.75 years (median, 31; SD, 17.68; range, 2 - 70 years). One hundred and two patients (75.55%) were aged 45 or less. The patients aged between 16 and 30 years were the most numerous with 44 (32.59%) cases (Figure 1). Pupils and students were the socio-professional category most often involved with 42 (31.11%) cases, followed by civil servants with 25% (N = 34).The road traffic accident(RTA) was by far the first mode of injury with 101 cases (74.81%) and most of the patients were pedestrians hit by a car in 44 cases (Table 2). The motorcycles were involved in 33% of the RTA. The assaults were the second mechanism of injury with 20 cases.

At admission, 97 (71.85%) patients had a GCS of 7 or 8 points, 41 (30.37%) had pupillary anomalies. Their mean arterial pressure was <90 mmHg in 47.5% and between 90 - 120 mmHg in 45% of the cases respectively. Twenty-four (17.77%) patients had hemoglobin level < 10 g/dl and 27 (20%) had electrolytes imbalance.

GCS: Glasgow coma scale; CT: computed tomography.

Twenty-seven (20%) patients had additional extracranial injury, 4 thoracic injuries and 10 limb fractures (Table 2, clinical profile).

A brain CT scan was done for 115 (85.18%) patients. Intracranial and intracerebral hemorrhages were found in 85 (62.98%) and skull fractures in 30 (22.22%) cases respectively (Table 2, CT findings).

All the patients had empirical and systematic application of some BTF guidelines and recommendations such as: elevation of head over bed at 30˚, nasal oxygenation to maintain oxygen saturation > 92%, maintenance of normal body temperature, and normovolemia by isotonic saline infusion, mannitol osmotherapy, indwelling urinary catheter, nasal gastric tube for precocious enteral nutrition, deep vein thrombosis, stress and decubitus ulcers prevention. Conversely, 65% (N = 88) of the patients were not intubated, though 57.77% (N = 78) were sedated and 31 (23%) had mechanical ventilation. Thirty-four (25%) patients underwent a neurosurgical operation for intracranial, intracerebral hematoma removal; depressed skull fracture or acute hydrocephalus (Table 2). The 27 patients with additional extracranial injuries also had specific management of these injuries.

The median follow up period was 42 months (range 12 to 72 months). The overall mortality was 32.59% with 44 deaths (GOS 1). Thirty-two over 44 (72.72%) of the deaths occurred in patients with a GCS 7 - 8 on admission. Thirty-six

(81.81%) deaths occurred within the first week of hospitalization. The mortality was higher in patients managed conservatively (77.27% of deaths, N = 34) than in those who were operated (22.73% of deaths, N = 10). Ninety-one (67.41%) patients survived, 74 (54.81%) of them had persisting neurologic disabilities (GOS 3 - 4) while only 17 (12.59%) had a good recovery (GOS 5). The persisting neurologic disabilities were as follows: motor deficits, 43 (31.85%); high mental functions (memory, speech, thought) disturbances, 25 (18.52%); and sphincter disturbances, 6 (04.44%) cases respectively (Figure 2, outcome). The length of hospitalization was ≤7 days in 60% (N = 81) of the cases. The outcome was influenced by the GCS at admission, pupillary anomalies, and the length of hospital stay; endotracheal intubation and surgery.

The patients’ mean age of this series was 32.75 years. Almost 76% of the patients were 45 years or less. The second and third decades were the most involved with 32.59% of the cases. These data are consistent with findings from most authors and reflect the fact that TBI in general and STBI in particular, primarily affect young active populations involved in activities at risk [1] - [6] . It is well established that TBI is the leading cause of death in peoples less than 45 years in developed countries.

The male predominance for TBI and STBI is unanimously reported from the literature suggesting that males are more exposed to activities at risk for TBI [1] [2] [4] [5] [11] - [16] . In the present series, males represented 85% of the patients which is similar to that reported by Kinyanjui and Kiwango et al., [1] [2] . Therefore, male gender was a risk factor for sustaining a severe TBI.

The road traffic accident was the first mechanism of injury and pedestrians were hit by a car most of the time. In Ethiopia, pedestrian accounted for 50% of RTA deaths [3] . Kinyanjui had reported that despite having the lowest rate of car- owner per inhabitant in the world, sub-Saharan Africa faces the highest mortality rates due to RTA per in habitant in the world. This has been labeled “the silent epidemic” [1] . This reflects the fact that in sub-Saharan Africa, cars and roads are generally of poor quality and that driving rules are violated. These findings are obvious in this series where almost 50% of RTA victims were pedestrians hit by cars. Depending on the series, the other modes of injury for severe TBI are falls, assaults, sport and domestic injuries and gunshots in different variations [4] [5] [6] [11] [12] [13] [14] [15] .

Pupils and students represented one-third of patients who sustained a STBI in this series. This reflects the fact that most patients were young adults; most of them were pedestrians hit by cars or were motorcycle riders. This is a realistic picture of Cameroon’s social demography and is consistent with data reported from other African countries [1] [2] [3] .

Forty-seven point five percent of the patients had at least one episode of arterial hypotension. This frequency is close to those reported by Chesnut et al. and Ouedraogo et al. (45.8% and 34.6% respectively) [18] [19] but is quite high compared to the frequency reported by Stochetti [20] . This high occurrence of hypotensive episodes can explain the very low proportion of good outcomes from this series. Arterial hypotension is a well-known factor for poor prognosis in STBI [1] [2] [9] [11] .

Seven (5%) patients had at least one episode of arterial hypertension. High levels of arterial pressure can overcome cerebral vascular autoregulation and therefore worsen post-traumatic cerebral edema.

Hypoxemia (SpO₂ < 90%) was present on admission in six (04.44%) patients. This rate was very low compared to the 45% and 57% reported by Chesnut and Stoccheti [18] [20] . The difference is explained by the fact that their series were pre-hospital phase studies. Hypoxemia causes secondary brain insult and is frequently associated with arterial hypotension, grieving the prognosis of STBI patients. Hypoxemia should be promptly corrected whenever noticed [2] [8] [16] [21] [22] [23] [24] [25] .

Seventy-two percent of the patients had a GCS score of 7 or 8 at admission, while those with GCS score of 3 - 4 represented only 10%. The low proportion of patients with low GCS scores at admission was probably due to high mortality on the field, therefore these patients did not reach the emergency department or the ICU. According to El-Gindi and Abdel Azeem [3] , the documented mortality rates of victims of RTA with TBI in Africa were as follows: 20% - 30% at the site of trauma, 7% - 20% during transportation, and 2% - 10% on hospital admission. We can therefore extrapolate that 29% - 60% of victims of RTA with STBI in Africa might not reach ICUs. In the series from Petroni et al. GSC 7 - 8 and 3 - 4 represented 41.9% and 38.5% of the cases respectively [11] . In studies from developed countries, STBI with low GCS (3 - 5) often outnumber those with high GCS (6 - 8) on admission [17] [21] [26] . This discrepancy with our results reflects the fact that STBI patients with low GCS are prone to die before reaching ICU in our setting.

Thirty percent (N = 41) of our patients presented with a pupillary abnormality. The presence of a pupillary anomaly in a STBI patient is a poor prognostic factor and it reported frequency is variable [4] [9] [11] [17] .

Twenty percent of the cases of this series had an associated extra cranial injury. The prevalence of extracranial injury in STBI patients could be as high as 77% [4] [9] .

Anemia is a deleterious factor for outcome of STBI [9] [21] [22] [23] [27] . For this reason, the hemoglobin level should be maintained above 9 g/dl. In the present series, the hemoglobin level was at some time < 10 g/dl in 37% of the patients.

Hypernatremia is another factor related with poor outcome in STBI and is reported in 16% - 40% of patients and it is associated with increased mortality [9] [28] . Hypernatremia was recorded in 10.52% of the cases in this series. The other electrolytes disturbances noticed were hyperchloremia and hyperkalemia.

CT scan was done in 85% of the cases of this series. It is not possible to do a CT scan in all patients with STBI because some of them are clinically very unstable and therefore not transportable to the CT scan facility [4] , Most findings were intracranial or intracerebral hemorrhages (62.98%) and depressed skull fractures (22.22%). In their study, Oddo et al. [21] report that 79% of patients in their series had traumatic SAH or intraventricular hemorrhage on admission CT scan. In the series from Rejeb et al. and Petroni et al. [9] [11] , CT findings were also dominated by intracerebral and intracranial hemorrhages.

The Brain trauma foundation and other entities have established guidelines for the management of TBI in general and STBI in particular [8] [11] [16] [17] [22] - [27] [29] [30] [31] [32] [33] . The main risk in patients with STBI is the occurrence of intracranial hypertension which develops in more than half of the cases. A refractory intracranial hypertension is responsible for most deaths in STBI. Intracranial pressure monitoring is necessary for managing intracranial hypertension. ICP monitoring is not available in our setting and STBI patients were managed empirically. The presence of an acute mass lesion worsens the prognosis of STBI patients. For patients with such lesions, prompt neurosurgical intervention is mandatory in order to avoid brain herniation. The presence of an acute mass lesion requiring surgical removal is a strong factor for poor outcome in patients with STBI. In this series, 25% of patients had intracranial lesions requiring surgery. But, there was no significant difference in terms of outcome between patients who needed surgery and those who did not. Kinyanjui reports the same percentage of STBI patients managed at ICU and who were operated [1] .

In this series, patients were not routinely intubated although more than half were sedated. Tracheal intubation was performed only when a patient showed signs of imminent death. This explains why intubation seemed deleterious to patients in this series. Severe TBI patients although managed at ICUs do not have routine tracheal intubation in most series from sub-Saharan Africa in spite of the recommendations. In this area, the proportion of STBI patients who benefit from intubation is most often lower than one third [13] [15] .

All patients of this series had mask oxygenation and this explains why 95.65% had SpO₂ > 90%, though 65% were not intubated. Correcting any hypoxemic episode in the course of STBI improves outcome [8] [16] [21] - [26] [29] [30] [31] [32] .

Mechanical ventilation is not systematic but may be required in the management of patients with STBI. In this series, only 35% of the patients had mechanical ventilation. It was always preceded with tracheal intubation and was always performed after neurological deterioration of the patients was observed. The frequency of mechanical ventilation in the management of STBI in sub-Saharan Africa remains very low compared to developed countries (5.3% - 37% vs 96.3%) [13] [15] .

The mortality of this series was 32.59% with 44 deaths (GOS 1). Ninety-one (67.41%) patients survived. Seventy-four (54.81%) had permanent neurologic disabilities (GOS 3 - 4) and only 17 (12.59%) recovered fully. The mortality of this series is quite low when it is compared to the mortality reported in other sub-Saharan Africa series for STBI managed at ICUs [1] [2] [13] [15] in which the mortality varies from 54% to 91%. The mortality of the present series was similar to that reported from industrialized countries and this low mortality in our setting can result from empirical application of some guidelines. Only 12.67% (17/135) of the patients recovered fully (GOS 5) and the remainder were disabled at some extent. Kiwango et al. reported 63.6% disability prevalence in STBI in their series [2] . This shows the unanimously admitted poor prognosis related with severe TBI. Although most of the survivors of this series were functionally independent (N = 70; 51.85%), 33.34% showed cognitive or behavioral impairment. In 2013 in Kenya, the prevalence of cognitive impairment in patients who had TBI varies from 32 to 50% depending on the scale used [1] . In long-term, although the majority of patients who sustained a STBI will show good physical recovery with independence in locomotion and basic life skills, most will remain with neuropsychological disabilities such as cognitive and behavioral disorders impeding on social reintegration [34] .

On multivariate analysis, the outcome was influenced by the length of hospital stay, pupillary abnormalities and intubation.