Pediatric Head Injury: A Study of 403 Cases in a Tertiary Care Hospital in a Developing Country (2024)

Abstract

Keywords: Glasgow Coma Scale, outcome, traumatic brain injury

INTRODUCTION

Severe traumatic brain injury (TBI) in children is Sa significant cause of morbidity and mortality worldwide.[1] Falls are the most common type of injury, followed by motor vehicle-related accidents.[2] Furthermore, child abuse remains a major cause of head trauma in children under 2 years of age. The percentage of each contributing factor differs between studies, and the distribution varies according to age group and sex. Infants and young children are more vulnerable to abuse because of their dependency on adults.[3] A force applied to the skull may be distributed evenly throughout the skull without causing a skull fracture (closed head injury) but damaging the less rigid brain tissue. When the skull vault is fractured, it is described as an open head injury. Open head injuries are much less common in children.[4] TBI is classified as mild (Glasgow Coma Scale [GCS] 13–15), moderate (GCS 9–12), or severe (GCS 3–8).[5] Children ≤10 years of age with a GCS of ≤8 or a strong suspicion of injury despite normal plain films (anteroposterior [A/P] and lateral views for children <10 years of age, without an A/P peg view), or if plain films are inadequate, should undergo computed tomography (CT) scanning of the cervical spine within an hour of presentation or when sufficiently stable.[6] Responsiveness is assessed with the Alert, Verbal, Pain, Unresponsive system and with the GCS[5] and its modified Paediatric GCS.[7]

CT is the most helpful and most definitive way to assess the severity of TBI. CT provides all essential information necessary to make a decision regarding the presence or absence of significant intracranial injury and whether emergency operative intervention is required.[8,9,10]

Although clinicians usually attempt to take a wide range of factors into account when making clinical decisions and assessing prognosis, there is probably a redundancy in this effort to be complete. In practice, relatively few features have been found to contain most of the prognostic information.[11,12,13] These include the patient age, clinical indices indicating the severity of brain injury (e.g., the depth and duration of coma and other neurological abnormalities), and the results of investigation and imaging studies, particularly intracranial pressure (ICP) and CT scanning, which disclose the nature of brain injury and its effects on intracranial dynamics.[14]

MATERIALS AND METHODS

Four hundred and three (403) patients with TBI, 18 years or less in age, managed in the Department of Neurosurgery, Sher-i-Kashmir Institute of Medical Sciences, Srinagar, over a period of 2 years from August 2011 to September 2013, were enrolled in the study.

The patients were evaluated on the basis of a predetermined pro forma. Detailed history of the patients was taken (including patients’ bio data, age, and mode of injury). Patients were subjected to detailed general physical examination, systemic examination, and central nervous system (CNS) examination including GCS and pupil size and reaction. Based on GCS, the patients were divided into mild head injury (GCS 13–15), moderate head injury (GCS 9–12), and severe head injury (GCS ≤ 8) categories All the patients were subjected to plain CT scan head, and CT findings were noted.

After the preliminary resuscitation and workup, the patients were managed conservatively or surgically as per indications. The outcomes of all these patients were assessed by Glasgow Outcome Scale and divided into good (normal, moderate disability) and poor (severe, vegetative, dead) outcome. Outcome was assessed in relation to age, sex, GCS, pupil size and reaction, CT scan features, intervention, and associated injuries.

RESULTS

A total of 403 patients aged between 1 day and 18 years with a mean age of 8.398 years with SD of ± 5.5228 were included in the study comprising 252 males (63.25%) and 151 females (37.75%), with a male-to-female ratio of 5:3. It was evident from our series that age (P = 0.478) and sex (P = 0.165) had no statistical significance in outcome [Table 1].

The most common modes of injury were fall 228 (56.6%) followed by road traffic accidents (RTAs) 138 (34.2%), assault 10 (2.5%), and others 27 (6.7%) which include sports injury, hit by some object on head, and firearm injury. RTAs had a poor outcome in 16.6% while patients with fall had a poor outcome of 7.8% (P = 0.052) [Table 1].

Majority of our patients had a GCS of 13–15 (mild head injury), 229 (57.3%) followed by 9–12 (moderate head injury) 119 (29.8%), followed by 8 or less (severe head injury) 52 (13%). In group of patients in the category of GCS ≤8, poor outcome was seen in 65.3%, followed by patients in group GCS 9–12 at 2.45% succeeded by group of patients with GCS 13–15 at 2.6%, which was statistically significant (P < 0.0001) [Table 1].

Out of the 403 patients, 354 (87.8%) patients had normal pupils, 37 (9.2%) had anisocoria, and 12 (3%) patients had fixed dilated pupils. Fixed dilated pupil had poor outcome (100%) followed by anisocoria (40.5%) and normal pupils (16%), which was statistically significant (P < 0.0001). CT scan findings were noted as normal in 86 patients (21.3%), isolated skull fracture in 89 (22%), contusion or hematoma in 98 (24.3%), extradural hemorrhage (EDH) in 66 (16.4%), subdural hemorrhage (SDH) in 35 (8.7%), pneumocephalus or aerocele in 13 (3.2%), brain edema in 8 (2%), and subarachanoid hemorrhage in 8 (2%). Among the mode of injury, it is evident that diffuse brain edema had poor outcome in 50%, SDH in 25.7%, contusion in 13%, while in EDH, it was 4.5% [Table 2]..

From our series, we inferred that poor outcome was associated with the highest (60%) being midline shift (MLS) >3 mm and it was 13.6% with MLS of <3 mm and it was 8.5% in patients with no MLS (P < 0.001) [Table 2]. Out of the 403 patients, 276 (69%) were managed conservatively and 124 (31%) patients were managed surgically. The various surgical procedures performed in patients include fracture debridement and elevation in 44 (35.48%), hematoma or contusion removal in 38 (30.64%), decompressive craniotomy in 9 (7.25%), hematoma or contusion removal with fracture debridement in 12 (9.67%), and hematoma removal with decompressive craniotomy in 21 (16.93%) [Table 3]. We followed up our patients for 10 days to 12 months, with a mean of 4.63 months. In the operated group, wound infections occurred in 22 (5.5%), cerebrospinal fluid (CSF) leak in 21 (5.3%), aspiration or hospital-acquired pneumonia in 15 (3.8%), CNS infections (meningitis or subdural abscess, etc.) in 6 (1.5%), and recollection of hematoma after surgery (SDH) in 6 (1.5%) cases.

While doing the survey for associated injuries, fractures of face were noted in 42 (10.5%), limb fractures in 27 (6.8%), abdominal trauma in 16 (4%), spinal trauma in 11 (2.8%), chest trauma in 9 (2.3%), multiple traumas in 12 (3.1%), and isolated head trauma in 286 (70.9%). It was noticed that chest, spinal, and multiple injuries were associated with a poor outcome (P < 0.0001) [Table 4].

DISCUSSION

Although in the developed world, with the advent of highly specialized Intensive Care Units and a high level of multidisciplinary approach, the outlook of the TBI has improved dramatically, it still continues to be a major challenge for the neurosurgery units in our part of the world. A pediatric TBI as is true in adults occurs due to a variety of reasons and generally has a good prognosis. These injuries more often than not are not isolated and are associated with polytrauma and are treated more or less on the same lines as the adults.[15,16,17,18]

There are discrepancies in the literature when defining the age point where prognosis significantly worsens. For example, there has been disagreement regarding the pediatric age group. One group of reports has indicated that outcome tends to be better in children under 10 years of age,[19,20] while others report that children under five have a higher mortality rate.[21] Although in our series there was no difference in poor outcome in children below 5 years or above 5 years as was reported by Suresh et al.,[22] there was slightly higher poor outcome above 12 years that was not statistically significant. The importance of age as a prognostic factor has been a subject of controversy. Luersson et al.[23] and Braakman et al.[11] have reported age as the strongest factor for mortality and morbidity. Although literature supports age as the stronger factor of mortality and morbidity in severe head trauma, these studies compare adults with children. In our series, we had compared only children and all grades of trauma were taken into consideration.

In our series, the most common modes of TBI were fall 228 (57%) with good outcome in 210/228 (92%) patients, followed by RTAs 135 (33.8%) with good outcome in 115/135 (83%), followed by assault in 10 (2.4%) with good outcome in 100%, followed by other modes in 27 (6.7%) with good outcome in 92.5%.

The patients with low GCS had a poor outcome as is expected. The patients who had a GCS of 13–15 (229 [57.3%]) had a poor outcome in 6 (2.6%), followed by GCS of 9–12 (119 [29.8%]) who had a poor outcome in 3 (2.52%), followed by GCS of 8 or less (52 [13%]) who had a poor outcome in 34 (65.4%). Suresh et al.[22] reported poor outcome in the group of GCS 3–5 as 58.5%, GCS 6–8 had 35.2%, GCS 9–12 had 11.4%, and GCS 13–15 had 1.3%. Beca et al.[24] and Kuday found that the initial GCS score was the single most important factor affecting outcome (P < 0.00001).[25] Astrand et al.[26] reported poor outcome in GCS 14–15 in 0%, in 9–13 as 6.2%, and < 8 had 22% poor outcome. Ong et al.[27] and Aldrich et al.[4] reported that low GCS did not always accurately predict the outcome in the absence of hypoxia or ischemia. In our series, we found a significant impact of GCS on outcome. The developing and underdeveloped areas like ours have poorer outcome due to lack of prehospital resuscitation and late presentation to hospital.

Out of the 403 patients, 351 (87.8%) patients had normal pupils, 37 (9.3%) had anisocoria, and 12 (3%) patients had fixed dilated pupils. Poor outcome in patients with normal pupils was 4.55%, patients with anisocoric pupils was 40.5%, and with fixed dilated pupils was 100%. Astrand et al.[26] reported 100% poor outcome in dilated pupils unresponsive to light. Suresh et al.[22] reported poor outcome of 49.3% in patients of abnormal pupils and 7.4% in normal pupils. Jennett et al.[12] reported poor outcome of 96% in fixed dilated and 50% in normal pupils. Francel et al.[28] stated that papillary response is not a good predictor of outcome. In our series, we found abnormal pupillary response being the strongest predictor of outcome.

Out of the 403 patients in our series, CT scan findings were noted as normal in 86 patients (21.3%) [Table 2], with good outcome in 71/86 (88%) and poor outcome in 11.6%. These results closely match those of Van Dongen et al.[29] in whose series patients with normal CT were 12% and good outcome was seen in 78% and poor outcome in 22%. Lobato et al.[30] reported normal CT in 10% patients. The extent of a skull fracture is proportional to the severity of brain injury which clearly does not apply to the pediatric age group.[28]

We found isolated skull fracture in 89 (22%) patients with good outcome in 97.8% and poor outcome in 2.2%. Suresh et al.[22] had 17% patients with isolated skull fracture with good outcome in 94.1% and poor outcome in 5.9%. The probability of associated intracranial hematoma with skull fracture in children is half of that of adults.[31]

Extradural hematoma is significantly less common in children than in adults and is even more rare in infants.[32] EDH can occur without fracture in children more commonly.[28] We had noticed EDH in 66 (16.4%) patients with good outcome in 95.5% and poor outcome in 4.5%. EDH reported by Suresh et al.[22] was at 28% with poor outcome in 8.4%. Astrand et al.[26] reported 48% of patients of EDH with good outcome in 98% and poor outcome in 2%. The mortality rate in children with EDH ranged from 7% to 15%, 5%–10% of patients also had residual neurological deficits.[28]

In our series, contusions/hematoma was seen in 98 (24.3%) with good outcome in 86.8% and poor outcome in 13.2%. Suresh et al.[22] reported 16.7% with poor outcome in 18.2% and Lobato et al.[30] reported that outcome was better in EDH. The outcome was unfavorable in patients with intracerebral hematomas and hemorrhagic contusions.[33]

SDH is seen six times more often in infants than in toddlers.[28] The outcome of patients with SDH is significantly worse than that of patients with EDH, mainly because of the underlying brain damage accompanying SDH and the resultant intracranial hypertension. In our series, SDH was seen in 35 (8.7%) patients and out of those poor outcome was noted in 9 (25.7%) patients. Suresh et al.[22] had recorded SDH in 10.33% of patients out of whom 35.3% had poor prognosis. Yi et al.[34] and Tomberg et al.[33] recorded 25% and 17.1% of patients having SDH and none of them had good prognosis. In our series, EDH was seen more common than SDH and possibly it is because of higher incidence of fractures in our series, i.e. 89 (22%) patients and EDH was mainly because of fracture line hematoma.

Diffuse brain swelling occurs in approximately 50% of children with severe head injury. The outcome is significantly better in children as compared to patients with operable mass lesion.[33] Diffuse swelling of brain may develop more readily in children because of the lack of CSF available for displacement. Children with CT scan indicating of diffuse axonal injury but without diffuse cerebral edema generally did not have sustained increased ICP and more than two-thirds attained a favorable outcome. Diffuse brain swelling with or without diffuse axonal injury demonstrated by the first CT scan was related to high mortality. In our series, diffuse brain edema was observed in 8 (2%) with poor outcome in 4 (50%) patients. Suresh et al.[22] reported the incidence of diffuse brain edema as 30% with poor outcome in 25%.

Quattrocchi et al.[35] found a prognostic significance of the presence or absence of MLS on the admission CT. Athiappan et al.[36] found the prognostic value of MLS to be more important in patients with single contusions or intracerebral hematoma than for those with multiple lesions and extra-axial or subdural hematoma. In our series, we found poor outcome in 8.5% and good outcome in 91.5% patients without MLS, patients with MLS <3 mm had poor and good outcome of 13.7% and 86%, respectively, and patients with MLS >3 mm had poor outcome and good outcome of 60% and 40%, respectively. The presence of MLS was associated with a poor outcome in 50% of cases, whereas the absence of MLS was associated with a poor outcome in only 14% of cases (P < 0.05).[35] Lobato et al.[30] reported that bad outcome can be predicted correctly as 68% when MLS >1.5 cm.

In our series, we found that 80% of patients had pure TBI, 10% had facial trauma, 7% had limb fractures, 4% had abdominal solid organ injury, 3% had spinal injuries, 3% had chest trauma, and 3% had multiple injuries. Jagannathan, et al.[1] reported orthopedic (50%), abdominal (27%), thoracic (20%), facial (20%), and multiple injuries (48%).  Paret et al.[37] reported chest trauma in 62%, limb fracture in 32%, facial fractures in 29%, abdominal solid organ lesions in 20%, spinal cord injuries in 5% and multiple in 67%. This difference is because we had taken all patients into consideration irrespective of grade of injury, while others have taken severe TBI patients into consideration.[1,37]

The overall outcome of our series of patients irrespective of their grade revealed a mortality of 5.7% and normal outcome of 81.9% with 4.7% patients being completely dependent for their day-to-day activities and 7.4% patients have moderate disability [Table 5].

REFERENCES