ORIGINAL PAPER

Conservative management of significant supratentorial epiduralhematomas in pediatric patients

Muhammad Babar Khan & Muhammad Riaz &

Gohar Javed

Received: 6 December 2013 /Accepted: 17 February 2014# Springer-Verlag Berlin Heidelberg 2014

AbstractBackground Radiologically significant epidural hematomasare generally treated with craniotomy and evacuation.However, pediatric patients with such hematomas may havenormal neurological examinations. We aim to report the pre-senting features, management, and outcomes of pediatric pa-tients who underwent conservative management of such he-matomas at our center and also show that serial neurologicalexaminations are a safe and effective way of managing thesepatients.Methods A retrospective review of pediatric patients withextradural hematoma (EDH) thickness of at least 1 cm andwho were conservatively managed was performed. All pa-tients were followed up in neurosurgery outpatient clinics afterdischarge at 1-month, 3-month, 6-month, and 1-year intervals.Detailed neurological examination was performed at eachvisit, and a standardized questionnaire was also used to doc-ument the parent’s subjective opinion of the patient’s qualityof life.Results A total of 17 patients satisfied the study criteria andwere included. Conservative management was successful in15 patients, while a craniotomy with evacuation of hematomahad to be performed in two patients. All patients had a GOSscore of 5 on 1-year follow-up, had normal schooling, andreported complete satisfaction with the management protocol.

Conclusion Conservative treatment is an optimal treatmentoption, and patients can be followed safely using a protocolof serial neurological examinations. A center must have re-sources to perform a craniotomy with evacuation of EDH incase of neurological worsening and be able to provide trainedstaff to carry out serial neurological examinations beforetreating these patients conservatively.

Keywords Epidural hematoma . Conservative management .

Pediatric patients . Craniotomy . Safety . Complications .

Treatment

Abbreviations used in manuscriptEDH Extradural hematomaER Emergency roomGCS Glasgow Coma ScaleGOS Glasgow Outcome ScoreICP Intracranial pressurePTSD Post traumatic stress disorderRTA Road traffic accident

Introduction

The incidence of epidural hematoma (EDH) is reported be-tween 1 and 6 % in pediatric patients with traumatic closedhead injury [1–3]. Traditionally, craniotomy with evacuationof hematoma has been accepted as the optimal managementstrategy for EDH [2, 4–6]. However, there is no uniformconsensus on the exact surgical indications [7–10]. The suc-cessful nonsurgical management of small asymptomatic EDHhas been well reported in literature [4, 8, 11–14]. Greater

M. B. Khan :M. Riaz :G. Javed (*)Section of Neurosurgery, Department of Surgery, Aga KhanUniversity Hospital, Karachi, Pakistane-mail: gohar.javed@aku.edu

M. B. Khane-mail: babarkhan08@gmail.com

M. Riaze-mail: muhammad.riyaz@aku.edu

Childs Nerv SystDOI 10.1007/s00381-014-2391-x

usage of computed tomography (CT) scan in patients withtraumatic head injury has resulted in an increase in the numberof patients with significant EDH of greater than 1 cm thick-ness on CT scans but normal neurological examination [6].These pediatric patients can potentially be managed conser-vatively, and so far, two case series comprising 13 and 6children have been reported in literature detailing the conser-vative management of such EDHs [5, 6].

The aim of this report is to describe the presenting features,management, and outcomes of conservative management ofsignificant EDH in 17 pediatric patients which makes this thelargest series of its kind in literature. Moreover, we used CTscan less frequently compared to earlier reports [5], and ourresults show that serial neurological examinations are a safeand effective way to monitor these patients.

Methods

This study is a retrospective reviewwhich included all patientsunder the age of 16 years managed conservatively for signif-icant EDH at our center between January 2001 and June 2012.We included patients with post resuscitation Glasgow ComaScale (GCS) scores of 15, clinically none, or mild to moderatesigns of elevated intracranial pressure (ICP) and EDH thick-ness of at least 1 cm. Exclusion criteria included patientswhose scans or hospital charts could not be retrieved, patientswith spontaneous EDH, and patients with EDH of unknownetiology. This led to a final inclusion of 17 patients in ourseries. The hospital and outpatient charts as well as radio-graphic studies of all patients were meticulously reviewed andall data extracted onto predesigned questionnaires, and adatabase was created. We included details of demographics,the time since injury and admission in emergency room (ER),the cause of injury, the presenting symptoms, GCS score,pupillary characteristics, cerebellar signs, facial nerve weak-ness, cervical tenderness, positive findings on neurologicalexamination, respiratory pattern, signs of skull base fracture,lucid interval, radiological findings, details of patient manage-ment, complications, and Glasgow Outcome Score (GOS) onlong term follow-up in our database.

All the patients were under close clinical observation insurgical special care units with facilities for one-to-one nurs-ing care for 48 h. Serial neurological examinations wereperformed hourly by a neurosurgery resident or resident nurseand consisted ofmonitoring of vitals including blood pressure,temperature, pulse, respiratory rate and pattern, GCS scoring,pupillary size and reaction to light, and power and reflexes inall limbs. Hemodynamic stabilization was ensured with IVfluids. The patients were then shifted to surgical wards whereserial neurological examinations were performed every 2 h.The patients had CT scans on admission, discharge, and incase of clinical deterioration. We have facilities to provide

immediate surgery within an hour to admitted patients, and alow threshold was maintained for surgical intervention whichwas performed if the child showed signs of local brain com-pression or herniation, any worsening neurological deficits,and if cardiorespiratory abnormalities were detected. We hadto eventually perform surgery in two patients who recoveredcompletely. All patients were followed up in neurosurgeryoutpatient clinics after discharge at 1-month, 3-month, 6-month, and 1-year intervals. Detailed neurological examina-tion was performed at each visit, and a standardized question-naire was also used to document the parent’s subjective opin-ion of the patient’s quality of life.

Results

A total of 17 patients were included in this study. This includ-ed 13male and 4 female patients. Themean age was 9.8 years.The mean duration from injury to assessment in our emergen-cy room (ER) was 29.5 h. Only three patients presented within24 h of trauma. Ten of these patients presented directly to ourcenter, while seven had been evaluated at other centers. Themechanism of injury included falls in seven patients, motorvehicle accidents in eight patients, and assault in two patients(Table 1). No patient had any focal neurological deficits. Allpatients had bilateral normal pupillary reactions to light. Nocerebellar signs were detected in any patient, and all patientshad normal breathing patterns. The presenting GCS was 15 inall patients. Sixteen patients had mild to moderate headaches.Eleven patients had one or more episodes of vomiting. Onepatient developed paresis of contralateral 6th nerve on day 6after injury which had resolved by day 15.

All hematomas were supratentorial and localized to theright side in eight and to the left side in nine patients. Themean midline shift associated with hematomas was 0.5 cm.The mean thickness of EDH in our series was 1.6 cm. Parietallobe was the most commonly affected, and the hematoma waspredominantly parietal in nine, frontal in six, and temporal intwo patients. All patients had patent basal cisterns, and none ofthe patients had hypo dense or mixed density hematomas oninitial scanning. Fractures of the skull were identified in 12patients, these crossed the middle meningeal artery in sevenand dural sinuses in five patients (Table 2).

The mean hospital stay was 11.2 days (Table 1). All thepatients were initially managed according to a conservativeprotocol; however, a craniotomy with evacuation of hemato-ma had to be performed in two patients in whom neurologicalworsening was noticed. One patient with a predominantlyfrontal hematoma developed photophobia and worseningheadaches on day 7 after injury (Fig. 1). The other patientwho was immediately operated upon also had predominantlyfrontal hematoma and showed improvement in symptoms byday 3; however, she had a recurrent severe persistent headache

Childs Nerv Syst

on day 6. The patients recovered completely from surgerywithout any residual deficits, and outcome was comparableto other conservatively managed patients.

On the one yearly follow-up visit, all patients had a GOS of5. None of the patients had developed worsening headaches orepilepsy or required any antiepileptic drugs. All patients hadreturned to the pre morbid social and intellectual levels. Oneof the patients whose cause of injury was assault displayedsymptoms of post traumatic stress disorder (PTSD) and wasbeing treated for it as well. All the parents showed satisfactionwith our treatment protocol of EDH (Table 3).

We noticed a decrease in falls and an increase in road trafficaccidents (RTAs) as the cause of EDH as the age increased in ourpatients.We tested for this correlationwith a Spearman’s test, andthe results were statistically significant (ρ=0.76, p value=0.002).

Discussion

Before CT scans, EDHs were diagnosed by cerebral angiog-raphy, pneumoencephalography, and exploratory burr holes[14]. Literature from that time shows that neurosurgeons werereluctant to use these invasive techniques in patients who weremildly symptomatic [14]. Patients in whom these invasiveinvestigative modalities were used often had serious neuro-logical dysfunction and were either comatose since the time oftrauma, or had persistent severe deficits or showed neurolog-ical decline after an initial lucid interval. Craniotomy withevacuation of hematoma was a life saving procedure in thesepatients [14]. A widely prevalent philosophy thus came intobeing that EDH must be immediately surgically evacuated ondiagnosis to prevent the inevitability of brainstem compres-sion [14]. The widespread availability and use of CT scanscoupled with a decrease in restrictions on the indications ofCT scanning have resulted in a greater proportion of patientswith mild symptoms being scanned [6]. Thus, EDH is increas-ingly being diagnosed in patients who would otherwise havenot received neurosurgical consultation. Although the in-grained fear of neurological worsening has meant that manypatients withmild symptoms and CTevidence of EDH are stillsurgically treated, an increasing body of literature hasemerged detailing the nonsurgical management EDH [5, 6,8, 11–16]. Servadei et al., while analyzing a large series of 158patients inclusive of all age groups from a multicentre study,concluded that patients with EDH thickness of less than 1 cmcan be managed conservatively [15]. Two smaller case serieshave even demonstrated successful conservative managementof EDH greater than 1-cm thickness in pediatric patients, andour series is the largest of its kind demonstrating successfulmanagement of EDH greater than 1 cm in thickness to the best

Table 1 Admission characteris-tics and management Patient characteristics Number

(%)Mean±standarddeviation

Gender Male 13 (76)

Female 4 (34)

Age (years) 9.8±6.3

Duration from injury toassessment (h)

29.5±47.9

Mechanism of injury Motor Vehicle accidents 8 (47)

Falls 7 (41)

Assaults 2 (12)

Admission GCS 15 17 (100)

>15 0 (0)

Presentation Directly to our center 10 (59)

Initially evaluated and managed atother centers

7 (41)

Hospital stay (days) 11.2±7.8

Management Successful nonsurgical 15 (88)

Conversion to surgical 2 (12)

Table 2 Radiological findings

Radiological Characteristics Number (%) Mean±standard deviation

Localization

Predominantly frontal 6 (35)

Predominantly parietal 9 (53)

Predominantly occipital 0 (0)

Predominantly temporal 2 (12)

Thickness of EDH (cm) 1.6±0.4

Midline shift (cm) 0.5±0.4

Clot density

Hyperdense 17 (100)

Hypodense/mixed density 0 (0)

Skull fractures

Present 12 (71)

Absent 5 (29)

Childs Nerv Syst

of our knowledge [5, 6]. It is important to recognize this poolof patients with only mild symptoms, no neurological deficits,and GCS score of 15 with a significant hematoma on CT scanas they can safely be managed nonsurgically.

Pang et al. noted that the outcome of medical managementdepended on the size, location, configuration, rapidity ofaccumulation, associated intradural lesions, extra cranial ex-travasation of blood through skull diastases, and also the ageof the patient [14]. Children have better outcomes with nonsurgical management primarily for two reasons. Firstly, theorigin of bleeding is venous mostly in pediatric patients, whileit is arterial in adults. Secondly, children are better able to copewith a rise in intracranial pressure due to a combination ofopen fontanelles, large extra cerebral spaces, and basal cis-terns as well as unfused suture lines [5, 6]. These factors meanthat conservative management is more likely to be successfulin children. Even in adults, EDHs with a thickness of less than1.5 cm, midline shift of less than 0.5 mm, and volume lessthan 30 ml have been successfully managed conservatively[8].

The mechanisms of injury in our series were quite differentfrom series previously reported [5, 6]. Jamous et al. andBalmer et al. had one patient with RTAs in their series of 6and 13 patients, respectively, while the predominant mecha-nism of injury in our series was caused by RTAs. In our series,the relationship between age and RTA as mechanism of injuryleading to EDH was also statistically significant. However,

this had no appreciable effect on outcome of conservativemanagement which was excellent and comparative in bothgroups.

About 70 % of our patients had associated skull fractures.Fractures closer to a major dural vessel have been significantlylinked to clinical deterioration in patients with EDH [13]. Thiswould imply a worse prognosis for such patients. However,Tuncer et al. found that the rate of resorption of hematoma issignificantly greater in patients with skull fractures and con-cluded that skull fractures should not be considered as riskfactors, but rather a sign of favorable prognosis of conserva-tive management [16].

Type A and Type B EDH have been described [14]. In typeA EDH, net expansion of the clot occurs approximately be-tween the last days of week 1 to week 3 post injury. Thesuccess of conservative treatment in these patients dependsupon the interplay between ICP bufferingmechanisms and therate and magnitude of EDH expansion [14]. Symptoms maythus transiently worsen in these patients before improving.However, we maintained a low threshold for surgery in suchcases where symptoms markedly worsened for fear of perma-nent neurological deficits. Further research to identify accu-rate predictors of success and failure in these patients isrequired. Conversely, type B EDH show net resorption with-out an expansile phase and can always be managed conserva-tively if the patient is able to tolerate the initial EDH.Complete resorption of EDH has been reported to occurbetween 21 to 90 days [14].

The outcomes in patients who are initially conservativelymanaged but later require craniotomy have traditionally beena subject of much concern. Two (9.5 %) of our patientseventually required craniotomy with evacuation of hematomaas their clinical symptoms worsened. None of the patients inJamous et al. series and one out of thirteen patients (7.6 %) inBalmer et al. series required a craniotomy with evacuation [5,6]. The outcomes in patients who require craniotomy andevacuation of EDH were similar to the patients who wereconservatively treated in our series and also in Balmer

Fig. 1 a Axial CT scan of a 12-year-old patient shows a right frontalEDH 2 cm in thickness. The scan shows normal gray and white matterdifferentiation without any midline shift. b Axial CT scan of the samepatient taken 2 days later shows no appreciable differences from the

earlier scan. c Axial CT scan of the same patient taken 3 months afterdischarge demonstrates marked resolution of the EDH. Some residualdural calcification can be appreciated

Table 3 Summary ofoutcomes on 1 yearfollow-up

No. of disabilities reported 17

New onset headaches 0

Epilepsy 0

Normal school performance 17

PTSD 1

Subjectively good quality of life 17

Visual/Auditory problems 0

Regular requirement of analgesics 0

Childs Nerv Syst

et al’s. series [5]. Knuckey et al. reported that the outcome ofpatients whose clinical condition deteriorates depends on theirinitial GCS scores [13]. All patients whose condition deterio-rated and who had a GCS score of 15 at presentation hadexcellent outcomes in that series [13]. Other authors havereported that the level of consciousness, pupillary status,decorticate or decerebrate posturing, respiratory, and cardio-vascular status pre operatively determine the post operativeoutcomes [14, 17]. Mortality rate increases with advancingstages of brain herniation [17]. However, outcomes are alwaysgood if early signs of brainstem compression are picked andan emergency evacuation of EDH performed [14]. Thus, theconservativemanagement of EDH is safe and effective as longas careful serial neurological examinations are performed andfacilities for rapid evacuation of EDH are available in case ofneurological worsening.

Previously, it has been reported that an EDH on a CT scanperformed till 6 to 8 h after trauma has an increased risk ofexpansion and neurological deterioration, while EDH diag-nosed after 8 h has a very little chance of deterioration and canbe managed conservatively [13, 18]. Our study supports thesefindings as all, except one, of our patients were scanned within8 h of trauma. Moreover, Knuckey et al. found an increasedmean age, a lower mean GCS score, and a lower mean EDHvolume in patients who subsequently deteriorated during con-servative therapy; however, the results were not statisticallysignificant [13]. Because only two patients out of a total 17required craniotomy with evacuation in our series, we couldnot perform any meaningful statistical analysis to correlateneurological worsening with age and volume. Similarly, as weincluded a highly select group of patients with presentingGCS scores of 15 in our series, we cannot comment on howGCS scores correlate with outcomes of conservative therapy.It must be borne in mind that this study reports on a relativelysmall number of selected patients, and neurosurgeons mustuse caution in applying our results to their patients. The key isindividualization of care to patient keeping in view the hospi-tal resources.

Conclusion

This study confirms in a larger series of patients that conser-vative treatment is an optimal treatment option, and volumealone is not an indication for surgery. It further shows thatpatients can be followed safely using a protocol of closeobservation employing serial neurological examinations.This can potentially reduce the costs and radiation exposureassociated with serial CT scan imaging. A center must haveresources to perform a craniotomy with evacuation of EDH incase of neurological worsening and be able to provide trained

staff to carry out serial neurological examinations beforetreating these patients conservatively.

References

1. Ciurea AV, Kapsalaki EZ, Coman TC, Roberts JL, Robinson JS 3rd,Tascu A, Brehar F, Fountas KN (2007) Supratentorial epidural he-matoma of traumatic etiology in infants. Child’s nervous system:ChNS. Off J International Soc Pediatr Neurosurg 23:335–341

2. Bullock MR, Chesnut R, Ghajar J, Gordon D, Hartl R, Newell DW,Servadei F, Walters BC,Wilberger JE (2006) Surgical management ofacute epidural hematomas. Neurosurgery 58:S7–S15, discussion Si-iv

3. Rocchi G, Caroli E, Raco A, Salvati M, Delfini R (2005) Traumaticepidural hematoma in children. J Child Neurol 20:569–572

4. Ciurea AV, Tascu A, Brehar FM, Nuteanu L, Rizea R (2009) A lifethreatening problem in infants: supratentorial epidural hematoma. JMed Life 2:191–195

5. Balmer B, Boltshauser E, Altermatt S, Gobet R (2006) Conservativemanagement of significant epidural haematomas in children. Child’snervous system: ChNS. Off J International Soc Pediatr Neurosurg 22:363–367

6. Jamous MA, Abdel Aziz H, Al Kaisy F, Eloqayli H, Azab M, Al-Jarrah M (2009) Conservative management of acute epidural hema-toma in a pediatric age group. Pediatr Neurosurg 45:181–184

7. Bejjani GK, Donahue DJ, Rusin J, Broemeling LD (1996)Radiological and clinical criteria for the management of epiduralhematomas in children. Pediatr Neurosurg 25:302–308

8. Chen TY, Wong CW, Chang CN, Lui TN, ChengWC, Tsai MD, LinTK (1993) The expectant treatment of "asymptomatic" supratentorialepidural hematomas. Neurosurgery 32:176–179, discussion 179

9. Chowdhury SN, Islam KM, Mahmood E, Hossain SK (2012)Extradural haematoma in children: surgical experiences and prospec-tive analysis of 170 cases. Turkish neurosurgery 22:39–43

10. Gerlach R, Dittrich S, Schneider W, Ackermann H, Seifert V,Kieslich M (2009) Traumatic epidural hematomas in children andadolescents: outcome analysis in 39 consecutive unselected cases.Pediatr Emerg Care 25:164–169

11. Tuncer R, Kazan S, Ucar T, Acikbas C, Saveren M (1993)Conservative management of epidural haematomas. Prospectivestudy of 15 cases. Acta Neurochir 121:48–52

12. Cucciniello B, Martellotta N, Nigro D, Citro E (1993) Conservativemanagement of extradural haematomas. Acta Neurochir 120:47–52

13. Knuckey NW, Gelbard S, Epstein MH (1989) The management of"asymptomatic" epidural hematomas. A prospective study. JNeurosurg 70:392–396

14. Pang D, Horton JA, Herron JM, Wilberger JE Jr, Vries JK (1983)Nonsurgical management of extradural hematomas in children. JNeurosurg 59:958–971

15. Servadei F, Faccani G, Roccella P, Seracchioli A, Godano U,Ghadirpour R, Naddeo M, Piazza G, Carrieri P, Taggi F et al(1989) Asymptomatic extradural haematomas. Results of a multicen-ter study of 158 cases in minor head injury. Acta Neurochir 96:39–45

16. Tuncer R, Acikbas C, Ucar T, Kazan S, Karasoy M, Saveren M(1997) Conservative management of extradural haematomas: effectsof skull fractures on resorption rate. Acta Neurochir 139:203–207

17. Weinman DF, Muttukumaru B (1968) The mortality from extraduralhaematoma. Aust N Z Acad Manag 38:104–107

18. Sullivan TP, Jarvik JG, Cohen WA (1999) Follow-up of conserva-tively managed epidural hematomas: implications for timing of re-peat CT. AJNR Am J Neuroradiol 20:107–113

Childs Nerv Syst