Intracranial Pressure Monitoring in Children with Severe Traumatic Brain Injury: A Retrospective Study.

INTRODUCTION: There is a paucity of literature on intracranial pressure (ICP) monitoring in children. The aim of this study was to ascertain whether ICP monitoring is useful in children with severe traumatic brain injury (TBI).
MATERIALS AND METHODS: Medical records of children between 1 and 12 years, admitted to neurocritical care unit with severe TBI in 2 years, were reviewed. The children were divided into two groups: study group (ICP monitored) and control group (ICP not monitored). Admission demographics, vital parameters, and computed tomographic scan findings were recorded. In the study group, date of ICP catheter insertion/removal with ICP values and treatment carried out for increased ICP were noted. Data on tracheostomy, duration of mechanical ventilation, hospital stay, and outcome at discharge were noted.
RESULTS: Demographic variables were comparable between the two groups. When adjusted for death, no significant difference was observed between the study and the control groups in median duration of mechanical ventilation: 35 days (95% confidence interval [CI]: 12-73) versus 55 days (95% CI: 29-55) (P = 0.96), hospital stay: 36 days (95% CI: 12-73) versus 58 days (95% CI: 29-58) (P = 0.96), and time to tracheostomy: 6 days (95% CI: 5-8) versus 5 days (95% CI: 4-7) (P = 0.49). Mortality rates, incidence of cranial surgeries, and outcome at discharge were also comparable.
CONCLUSION: ICP monitoring did not reduce the incidence of death, cranial surgeries, duration of mechanical ventilation, hospital stay, or improve the outcome at discharge in children with severe TBI.

INTRODUCTION

Increased intracranial pressure (ICP) is a major consequence of moderate and severe traumatic brain injury (TBI). Nearly 50% of comatose patients with TBI with abnormal computed tomography (CT) scan have high ICP.[1] Elevated ICP has many harmful effects on brain such as brain herniation.[2] In children with severe TBI, a normal CT of head does not exclude intracranial hypertension (ICH); therefore, it is recommended that the need for ICP monitoring should be determined by the depth of coma, in addition to radiographic imaging.[3] In this group of patients, the treatment of ICH may be initiated when the ICP exceeds 15mm Hg in infants, 18mm Hg in children up to 8 years, and 20mm Hg in older children.[45] The measurement of ICP is an invasive procedure and may be complicated by intracranial hemorrhage and infection. There is no uniformity in practice in this field after Chesnut et al.[6] reported that care focused on maintaining ICP less than 20mm Hg in adult patients with severe TBI was not superior to care based on imaging and clinical examination. It is not known whether this observation can be extrapolated to children. Not many centers practice ICP monitoring in children because of technical and financial reasons. ICP is routinely monitored in our institute in patients of all age groups with severe head injury. There is a paucity of literature regarding the use of ICP monitoring in children and its impact on outcome. Hence, this retrospective study was planned to assess the effect of ICP monitoring in children with severe TBI on outcome at discharge from the hospital.

MATERIALS AND METHODS

After obtaining approval from the institutional ethics committee for conducting this retrospective study, available medical records of children between 1 and 12 years, who were admitted to the intensive care unit (ICU) with severe head injury (Glasgow Coma Scale [GCS] score of less than 9) during a period of 2 years (June 2011 to May 2013), at our neurotrauma center were analyzed. The children were divided into two groups: study group in which ICP was monitored and control group in which ICP was not monitored. Monitoring was carried out by inserting a Codman ICP catheter into brain parenchyma. The primary outcome was divided into three categories (A = death, B = disability/persistent vegetative state, and C = no disability at discharge from hospital). The secondary outcome was to find out the correlation between observed variables and outcome at discharge and the adverse events encountered in both the groups.

Demographic data at admission such as age, sex, approximate body weight, and vital parameters such as heart rate (HR), blood pressure (BP), pulse oximetry (SpO2 [saturation of arterial oxygenation]), respiratory rate (RR), and GCS scores were noted. Findings of CT scan and Marshall Scale score[7] were also recorded. On the basis of the CT scan findings, diagnoses were categorized into four types: 1 = extradural hematoma, 2 = subdural/intraparenchymal/intraventricular hematoma, 3 = diffuse axonal injury/diffuse brain edema, and 4 = others, including hydrocephalus/ventriculitis/meningitis.

During ICU stay, all patients were mechanically ventilated to maintain a partial pressure of arterial carbon dioxide (PaCO2) of 35–40mm Hg and were sedated with a titrated infusion of midazolam and fentanyl. Variables recorded in the ICU were hemodynamic parameters, any episodes of hypoxia, hyperthermia, induced hypothermia, hypo- and hyperglycemia, coagulopathy, leukocytosis, thrombocytopenia, electrolyte imbalance, chest infection (documented chest X-ray findings) along with surgical intervention, and intraoperative course. Ventilator-associated pneumonia (VAP) was defined as a new-onset pneumonia in a patient after 48h of tracheal intubation.[8] The diagnosis of pneumonia was made if the patient had signs such as temperature exceeding 38.5°C, tachypnea (RR exceeding 40 breaths per minute), increased oxygen requirement, total leukocyte count (TLC) more than 15 × 103 cells/mm3, cultured pathogen from tracheal aspirate, together with a positive gram stain, plus an infiltrate on chest radiograph persisting for 48h or more.[9]

In the ICP-monitored group, duration of ICP monitoring with ICP values recorded at various time intervals, management strategies instituted for ICH such as administration of mannitol and/or hypertonic saline, requirement of sedative drugs, hyperventilation, cerebrospinal fluid (CSF) drainage by external ventricular drain, and decompressive craniectomy were noted. The treatment measures instituted for raised ICP were based on clinical features such as unequal pupils, bradycardia with hypertension, and fall in GCS score, as noticed by the attending physician. The duration of mechanical ventilation, tracheostomy, ICU, and hospital stay was noted. Appropriate statistical analysis was carried out. Quantitative variables such as hemodynamic parameters were analyzed using the Student’s t-test. Qualitative variables, such as sex distribution, were analyzed using chi-square test. Survival probability was calculated by log-rank sum test and Kaplan–Meier analysis. A P value less than 0.05 was considered statistically significant.

RESULTS

Medical records of 61 children with severe TBI who fulfilled the study criteria were analyzed. A total of 23 children underwent ICP monitoring (study group) and 38 children were not monitored for ICP (control group). GCS scores were dichotomized into two subcategories for analysis (I: GCS score of 3–5, II: GCS score of 6–8).

The demographic variables were comparable between the two groups [Table 1]. The two groups were also comparable with regard to predefined diagnoses (P = 0.26), GCS subcategories (P = 0.68), Marshall Scale score (P = 0.13), associated cervical spine injuries (P = 0.65), and the presence of neurologic deficits (P = 0.99) [Table 1]. No significant difference was observed in admission vital parameters such as HR, systolic blood pressure (SBP), diastolic blood pressure (DBP), RR, and SpO2 between groups [Table 2]. Various measures instituted for the treatment of ICH (such as head elevation, hyperventilation, osmotic diuretics, and decompressive craniectomy) were comparable between the two groups [Table 3]. No significant difference was reported in various investigations carried out at admission or at discharge in the two groups [Table 4].

Table 1

Comparison of demographic variables

Variables	Study group (n = 23)	Control group (n = 38)	P value
Age (years)	5 ± 2	5 ± 3	0.34
Sex (M/F)	15/8	27/11	0.85
Weight (kg)	10 ± 6	12 ± 8	0.73
Diagnosis category 1/2/3/4	1/19/3/0	4/23/8/3	0.26
Associated cervical injury (Y/N)	2/21	1/37	0.65
Neurologic deficits (Y/N)	5/18	7/31	0.99
GCS subcategory I/II	8/15	10/28	0.68
Marshall score (1/2/3/4/5)	1/6/4/1/11	9/7/2/5/15	0.13

M = male, F = female, Y = yes, N = no

Diagnosis category: 1 = extradural hematomas, 2 = subdural/ intraparenchymal/intraventricular hematomas, 3 = diffuse axonal injury/diffuse brain edema, 4 = others (hydrocephalus/ventriculitis/meningitis/etc.)

GCS subcategory: I = GCS score of 3–5, II = GCS score of 6–8

Table 2

Vital parameters (mean ± SD) at ICU admission recorded in the two groups, correlated with ICP monitoring (Y/N), diagnosis categories 1/2/3/4, and GCS subcategories I/II

Vitals	Study group (n = 23)	Control group (n = 38)	P value
Heart rate	108 ± 25	111 ± 33	0.72
Systolic blood pressure	102 ± 17	103 ± 15	0.58
Diastolic blood pressure	63 ± 12	62 ± 12	0.95
Respiratory rate	24 ± 8	22 ± 8	0.49
SpO2	97 ± 11	97 ± 16	0.47

SD = standard deviation, Y = yes, N = no

Diagnosis category: 1 = extradural hematoma, 2 = subdural/intraparenchymal/intraventricular hematoma, 3 = diffuse axonal injury/diffuse brain edema, 4 = others (hydrocephalus/ventriculitis/meningitis/etc.)

GCS subcategory: I = GCS score of 3–5, II = GCS score of 6–8

Table 3

Treatment measures instituted for ICH

Treatment	Study group (n = 23)	Control group (n = 38)	P value
Head elevation (Y/N)	21/2	35/3	0.45
Osmotic agents (Y/N)	23/0	37/1	0.67
Hyperventilation (Y/N)	10/13	15/23	0.28
Sedation (Y/N)	18/5	30/8	0.72
CSF drainage (Y/N)	13/10	16/22	0.56
Decompressive craniectomy (Y/N)	12/11	15/23	0.34

Y = yes, N = no

Table 4

Blood investigations at admission and discharge/death (mean ± SD)

Investigations	Admission			Discharge/Death
	Study group (n = 23)	Control group (n = 38)	P value	Study group (n = 23)	Control group (n = 38)	P value
Hemoglobin (gm/dl)	9.4 ± 3.8	10.2 ± 4.1	0.39	8.7 ± 3.2	9.5 ± 2.7	0.16
TLC (per cubic millimeter)	14,691 ± 6,951	12,080 ± 4,128	0.19	12,073 ± 3,126	10,938 ± 2,893	0.19
Platelet count (per cubic millimeter)	278 ± 153	267 ± 177	0.69	282 ± 128	273 ± 138	0.69
INR	1.36 ± 0.33	1.35 ± 0.54	0.38	1.20 ± 0.19	1.22 ± 0.26	0.98
Serum urea (mg/dl)	17 ± 6	23 ± 29	0.90	16.78 ± 3.87	17.34 ± 7.59	0.75
Serum creatinine mg/dl	0.37 ± 0.24	0.48 ± 0.50	0.27	0.31 ± 0.13	0.38 ± 0.20	0.18

SD = standard deviation, TLC = total leukocyte count, INR = international normalized ratio

Complications were noted in six children in the study group and seven children in the control group (P = 0.92 and P = 0.65, respectively). One child, in ICP-monitored group, who developed VAP, brain abscess, pneumocephalus, and hydrocephalus was subjected to barbiturate coma therapy, and the child died on the 73rd day of admission to the ICU. Coagulopathy was seen in three children in ICP-monitored group and two children in the control group. In the ICP-monitored group, two children required inotropic support to maintain BP; one child developed meningitis with septic shock and another had severe sepsis. No incidence of ICP-catheter-related infections was observed.

Incidence of cranial surgery (including decompressive craniectomy), VAP, readmission to ICU after discharge, and outcome at discharge were comparable between the two groups [Table 5]. The median time to tracheostomy was also comparable between the two groups (P = 0.50) [Table 5]. Within the ICP-monitored group, the children with GCS score of 3–5 had higher mean ICP than those with GCS score of 6–8: 28.6mm Hg (95% confidence interval [CI]: 11.9–45.3) versus 15.1mm Hg (95% CI: 11.3–18.9) (P = 0.003). The need for further cranial surgery was also higher in children with higher opening ICP: 25.8mm Hg (95% CI: 16.9–34.7, n = 14) than in children with lower opening ICP pressure: 10.6mm Hg (95% CI: 8.6–12.5, n = 9) (P = 0.0004). Eight children who died (Outcome A) had significantly higher mean ICP of 26.1mm Hg (95% CI: 13.6–38.7) compared to those who had only disability (Outcome B, 11 children) with ICP of 18.3mm Hg (95% CI: 8.4–28.2) or those who had good outcome at discharge (Outcome C, four children) with ICP of 11.5mm Hg (95% CI: 6.1–16.5) (P = 0.006).

Table 5

Observed and outcome variables during hospital stay and at discharge

Variables	Study group (n = 23)	Control group (n = 38)	P value
Cranial surgery	14 (61%)	17 (45%)	0.34
VAP	9 (39%)	7 (18%)	0.14
Median days to tracheostomy	6 (95% CI: 5–8)	5 (95% CI: 4–7)	0.50
Readmission	6 (26%)	3 (8%)	0.16
Outcome at discharge A/B/C	8/11/4	12/14/12	0.45
Death	8 (35%)	12 (32%)	0.65

VAP = ventilator associated pneumonia, Outcome at discharge: A = death, B = disability/persistent vegetative state, C = no disability

On univariate analysis, the overall outcome at discharge (n = 61) was significantly different with regard to GCS subcategories I and II, age, and diagnosis categories [Table 6]. Eleven (61%) children with GCS score of 3–5 died as compared to nine (21%) children with GCS score of 6–8 (P = 0.008) [Figure 1]. The outcome was also affected by age; children with lower age had poorer outcome (P = 0.008) [Figure 2] [Table 6]. Only three (60%) children in GCS subcategory I had favorable outcome as compared to nine (21%) children in category II. Twenty (47.6%) children in GCS subcategory II died, with no incidence of deaths in other diagnosis categories (P = 0.007). Outcome at discharge was not different with regard to the incidence of cranial surgery (P = 0.24) or the incidence of VAP (P = 0.25).

Table 6

Variables associated with outcome at discharge

Variables	Outcome A	Outcome B	Outcome C	P value
Age (median years)	3	4	6.5	0.007
Sex (M/F)	14/6	16/9	12/4	0.752
Diagnosis categories (1/2/3/4)	0/20/0/0	2/13/7/3	3/9/4/0	0.007
Associated cervical injury (Y/N)	0/20	2/23	1/15	0.449
ICP (Y/N)	8/12	11/14	4/12	0.457
VAP (Y/N)	5/15	9/16	2/14	0.246
Marshall score (1/2/3/4/5)	0/4/1/3/12	6/6/3/2/8	4/3/2/1/6	0.363
GCS (I/II)	11/9	5/20	2/14	0.008
Cranial surgery (Y/N)	13/7	12/13	6/10	0.243

Bold values indicate statistical significance

M = male, F = female, Y = yes, N = no

Outcome at discharge: A = death, B = disability/persistent vegetative state, C = no disability

Diagnosis category: 1 = extradural hematoma, 2 = subdural/intraparenchymal/intraventricular hematoma, 3 = diffuse axonal injury/diffuse brain edema, 4 = others (hydrocephalus/ventriculitis/meningitis/etc.)

Figure 1

Bar chart showing distribution of outcome at discharge of patients with GCS. GCS I = GCS score of 3–5, GCS II = GCS score of 6–8. Outcome at discharge: A = death, B = disability/persistent vegetative state, and C = no disability

Figure 2

Comparison of age with outcome at discharge Outcome at discharge: A = death, B = disability/persistent vegetative state, and C= no disability

On multivariate analysis, age (P = 0.013) and GCS subcategory (P = 0.010) were significantly associated with poor outcome. The duration of mechanical ventilation was comparable between the two groups when adjusted for death: 39 days (95% CI: 21–56) in ICP-monitored group and 42 days (95% CI: 23–61) in the control group (P = 0.96) [Figure 3A]. Similarly, duration of hospital stay was also similar between the two groups when adjusted for death: 42 days (95% CI: 23–61) in the study group and 49 days (95% CI: 27–71) in the control group (P = 0.960) [Figure 3B]. Among all children (n = 61), irrespective of ICP monitoring, the mean duration of mechanical ventilation was significantly different with regard to GCS subcategories: 14 days in GCS subcategory I (3–5) compared to 53 days in GCS subcategory II (6–8), P < 0.0001 [Figure 4A], and the presence of VAP (51 days with VAP, 40 days without VAP, P = 0.03) [Figure 4B]. The duration of mechanical ventilation was not affected by age (P = 0.20), diagnosis categories (P = 0.46), coexisting cervical injury (P = 0.22), or incidence of cranial surgeries (P = 0.64). On Cox proportional hazards regression analysis, GCS score of 3–5 was associated with a relative risk of 6.2 (95% CI: 2.5–15.7) of death and a shorter duration of ventilation. The duration of hospital stay in these children was significantly different with respect to GCS subcategories (16 days with GCS subcategory I [3-5] and 58 days with GCS subcategory II [6-8], P < 0.0001) [Figure 4C]. The duration of hospital stay was not affected by age (P = 0.14), diagnosis categories (P = 0.36), presence of VAP (P = 0.06), cervical injury (P = 0.25), or incidence of cranial surgeries (P = 0.60).

Figure 3

(A) Days of mechanical ventilation adjusted for death. (B) Days of hospital stay adjusted for death in both the groups

Figure 4

(A) Days of mechanical ventilation with regard to GCS subcategories (GCS I = GCS score of 3–5, GCS II = GCS score of 6–8). (B) Days of mechanical ventilation with regard to VAP (Y = yes/N = no). (C) Days of hospital stay with regard to GCS subcategories (GCS I = GCS score of 3–5, GCS II = GCS score of 6–8)

DISCUSSION

Effect of ICP on outcome

This study suggests that ICP monitoring can be safely carried out in children with severe TBI. However, the monitoring modality did not affect the outcome at discharge. Previous studies reported lower mortality as well as better functional outcome among children between 2 and 12 years with ICH when ICP was adequately controlled.[3] In a study by Alkhoury and Kyriakides,[10] ICP monitoring was found to be associated with a reduction in mortality in children with GCS score of 3. However, children in the ICP-monitored group had a significant longer hospital and ICU stay and more ventilator days. In our study, ICP monitoring did not alter duration of mechanical ventilation or hospital stay. The children who had mortality in this study had a significantly shorter length of hospital stay and duration of mechanical ventilation, reflecting the severity of their disease. Children with intermediate outcome (B) with disability/persistent vegetative state required longest period of ventilation and maximum number of days of hospital stay. Children with good outcome in category C had duration of mechanical ventilation and hospital stay intermediate between these two categories. They survived the initial insult and were discharged with minimum or no disabilities [Table 7].

Table 7

Outcome at discharge compared with death-adjusted mean duration of hospital stay and mechanical ventilation (days)

Outcome at discharge	Hospital stay	Mechanical ventilation
Death (A)	14 ± 5	14 ± 5
Disability/persistent vegetative state (B)	83 ± 8	74 ± 7
No disability (C)	43 ± 0	25 ± 0

A limited evidence is available to support ICP monitoring and aggressive treatment of ICH in children. In adults with severe TBI, Chesnut et al.[6] observed that care focused on maintaining monitored ICP at 20mm Hg or less was not superior to care based on imaging and clinical examination. This was a multicentric trial with the possibility that different management protocols were followed at individual institution.

Despite lack of convincing evidence, ICP monitoring has become the standard of care for children with severe TBI who are at risk for ICH. This can be attributed to the facts that the placement of an ICP catheter is associated with only a few risks and because the normalization of ICP is believed to play a role in maintaining cerebral perfusion pressure (CPP), optimizing oxygen delivery, and preventing cerebral herniation.[11] Bennett et al.[12] found that hospitals with higher patient turnover and with more frequent ICP monitoring had better patient outcome. Although the optimum pediatric ICP and CPP are yet to be determined, an ICP >20mm Hg is generally accepted as a threshold for the institution of treatment,[4] whereas a CPP of 50–65mm Hg is currently considered optimal. Brief increases in ICP that return to normal in 5min may be insignificant; however, sustained increases of >20mm Hg for more than 5min probably warrants treatment.[13] A similar protocol was followed at our institute. In hospitals where continuous CSF diversion procedure is used as a therapy, intraventricular catheter may be useful for accomplishing monitoring and therapeutic goals.[14] In our institute, the intraparenchymal Codman sensor transducing system is used for ICP measurement, which does not allow CSF drainage.

Effect of GCS subcategories on outcome

Despite extensive experience and research focused on TBI, the determinants of outcome in children remain unclear. Most of the pediatric outcome studies are focused on the variables recorded at admission in the emergency department. Even though admission GCS score is useful, GCS score obtained in the immediate post-traumatic period has not been proven to be a reliable predictor of long-term outcome.[6] A GCS score ≤5 has been found to be a significant risk factor for death in children after severe head injury.[15] In our study, children with GCS score of 3–5 had poor outcome compared to those children with higher GCS score of 6–8, despite undergoing similar number of cranial surgeries. Furthermore, GCS score of 3–5 was associated with shorter duration of mechanical ventilation (13 days) and hospital stay (16 days), but was associated with 6.2 times higher relative risk of death. Within the ICP-monitored group, in this study, children with GCS score of 3–5 had higher ICP than those with GCS score of 6–8 [Figure 5]. Children with higher ICP underwent greater number of cranial surgeries than those with lower ICP. Also, children with higher ICP had poorer outcome as compared to those with lower ICP values in the study group. This reflects the impact of the severity of injury as assessed by GCS score on the outcome of these children. Our findings are similar to those by Grinkevici ūte et al.[15] who also observed that GCS score of 3–5 had a very high sensitivity for predicting death.

Figure 5

ICP with respect to GCS subcategories (CAT). GCS I = GCS score of 3–5, GCS II = GCS score of 6–8, GCS CAT = Glasgow Coma Scale subcategory

Effect of age on outcome

In our study, age affected the outcome at discharge (P = 0.007). Younger children (age ≤3 years) were more likely to have poor outcome (mortality). Older children (>6.5 years) had more chance of favorable outcome [Table 6]. Normal values for mean arterial BP and hence, CPP are lower in children <3 years age. It has also been shown that children <3 years have less autoregulatory reserve than older children. This suggests that a lower ICP therapeutic target is more appropriate for infants and young children with TBI. In the past, management of TBI has used hyperventilation to lower ICP. Finally, in light of the varied pathophysiology of pediatric TBI, ICP management may need to be individualized in many cases.[4]

Effect of diagnosis on outcome

The outcome at discharge was also influenced by diagnoses. Children with extradural hematomas had better outcome, whereas those with subdural hematomas, intraparenchymal contusions, and intraventricular hemorrhages had worst outcome, with a mortality rate of 47.6%. Children with diffuse axonal injury, diffuse brain edema and hydrocephalus, ventriculitis, and meningitis had intermediate outcome. This signifies the importance of the type of injury. The highest intracranial insult was present in diagnosis category 2; hence, the incidence of mortality was also greater in that category. However, category 2 reflects a combination of subdural hematomas, intracranial contusions, and intraventricular hemorrhages, and mortality figures may be different in each category of trauma. Differentiation of these categories is usually not possible in severe TBI as there is a lot of overlap in the diagnosis. However, the diagnosis did not change the duration of mechanical ventilation or hospital stay.

Effect of VAP on outcome

The incidence of VAP was also similar in both the groups in this study. Pediatric patients who develop VAP usually have a longer duration of hospital stay compared to those without VAP.[16] In our study, the duration of mechanical ventilation was longer (51 days) in children with VAP than those without VAP (40 days). However, duration of hospital stay was not affected by VAP. Early enteral feeding and gastric ulcer prophylaxis may reduce the incidence of VAP. Similar institutional protocols were followed in this study. No incidences of ICP-catheter-related infections were reported in the study group, which suggests that the ICP monitoring, per se, did not add to infectious complications and was safe in this regard. Similarly, the incidence of VAP was not a determinant of outcome at hospital discharge in this study.

Effect of other variables on outcome

Normal BP and adequate oxygenation have been proposed as factors conferring a higher probability of survival following TBI. Conversely, hypotension, hypoxia, and aggressive, prophylactic hyperventilation may have detrimental effects on outcome. However, the time-dependent effects of these variables on the pediatric head-injured patient are unknown.[17] White et al.[18] found that patients with higher GCS scores at 6h after injury were more likely to survive. They found that survival, when adjusted for severity of injury, was associated with maximum SBP >135mm Hg, suggesting that supranormal BPs may be associated with improved outcome. In our study, admission vital parameters (HR, SBP, DBP, SpO2, and RR) and baseline investigations were comparable in both the groups. They were also comparable when correlated to GCS scores and admission diagnosis.

In addition to the primary injury related to trauma, secondary events (such as edema, inflammation, and vascular injury) can also play an important role in the pathophysiology of ICP elevation, thus complicating its management. Treatment strategies for raised ICP rely on balancing volumes of blood and CSF following the principles of the Monro–Kellie doctrine. This balance can be achieved using medications (such as sedatives and muscle relaxants), diverting CSF flow (by ventriculostomy), or allowing the expansion of the brain (via a decompressive craniectomy).[19] Mannitol administration has been associated with prolonged length of stay in ICU, without conferring survival advantage.[20] In our study, the administration of osmotic diuretics and use of other ICP-reducing measures (such as head elevation, sedation, and hyperventilation) were similar in both the groups; it implies that no impact was present on the outcome by use of osmotic agents or other ICP-reducing measures [Table 3]. However, further studies are required in this regard.

The median time to tracheostomy in both groups shows a trend toward early tracheostomy (<10 days), which reflects our institutional protocol and has been associated with better outcome in previous studies.[2122] This may also be a factor in the outcome at discharge, which is similar in both the groups.

Limitations

This study has a few limitations. It was carried out in a single tertiary-level hospital, which receives many referred cases. Hence, patients presenting to the hospital tend to have a greater severity of injury with possibly poor outcome. This may have led to selection bias. The study also has the usual limitations inherent to all retrospective studies. Furthermore, the long-term clinical outcomes, including cognitive function and quality of life were not studied. Larger prospective studies with long-term follow-up are needed to know the benefit of ICP monitoring in children for widespread clinical application.

CONCLUSION

In this study, ICP monitoring was safely conducted in children with severe TBI. However, monitoring did not reduce the incidence of death, cranial surgeries, duration of mechanical ventilation, hospital stay, or improved the outcome at discharge when compared to children who were managed without ICP monitoring. The GCS score and age at admission were main predictors of outcome at discharge.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.