Risk Stratification in Patients with Severe Traumatic Acute Subdural Hematoma

Article information

Nerve. 2017;3(2):50-57
Publication date (electronic) : 2017 September 18
doi : https://doi.org/10.21129/nerve.2017.3.2.50
Departments of Neurosurgery, Konkuk University Medical Center, Konkuk University School of Medicine, Seoul, Republic of Korea
Corresponding author: Sang Woo Song, Departments of Neurosurgery, Konkuk University Medical Center, Konkuk University School of Medicine, 120-1, Neungdong-ro, Gwangjin-gu, Seoul 05030, Republic of Korea, Tel: +82-2-2030-7357, Fax: +82-2-2030-7359, E-mail: his4u2@hanmail.net
Received 2017 August 29; Revised 2017 September 14; Accepted 2017 September 18.

Abstract

Objective

The aim of this study was to investigate risk grouping for surgical outcome in patients with severe traumatic acute subdural hematoma (ASDH).

Methods

Seventy-five patients showing low Glasgow Coma Scale (GCS) 3 to 8 were enrolled in this retrospective study. Clinico-radiologic findings were retrieved from electronic medical record and computed tomography. Prognostic factors from univariate and multivariate statistical methodology were included in a recursive partitioning analysis for risk stratification.

Results

One month after surgery, 54 patients (72%) had poor Glasgow Outcome Scale (GOS) 1 to 2 (unfavorable outcome). The surgical outcomes were stratified into three homogenous risk groups according to preoperative GCS and presence of basal cistern obliteration. The rate of favorable outcome and mortality significantly differ between the groups: 4.9% and 68.3% in patients with GCS 3 to 5, 23.1% and 53.8% in patients with GCS 6 to 8 and basal cistern obliteration, and 76.2% and 0% in patients with GCS 6 to 8 and without basal cistern obliteration.

Conclusion

The surgical outcomes of severe traumatic ASDH patients could be stratified preoperative GCS score and the presence of basal cistern obliteration. It is expected that this model will not only provide objective information when we make decisions about treatment, but it can also be a useful tool when discussing the patient’s prognosis with the patient’s caregivers.

INTRODUCTION

Acute subdural hematoma (ASDH) is a major clinical entity in traumatic brain injury (TBI) and is diagnosed by computed tomography (CT) as a hyperdense, extra-axial, crescent lesion between the dura and brain parenchyma1). The acceleration-deceleration force applied during traumatic insults can result in stretching and tearing of the cross-linking blood vessels and cortical arteries, which is the source of hematoma formation25). The pressure imposed on the cerebral tissue by a hematoma is not the only factor that affects neurological outcomes, because ASDH is commonly associated with brain edema, contusions, and diffuse axonal injury (DAI)1,6,7). This can increase midline shift (MLS) and make hematoma removal inappropriate for satisfactory patient outcomes8,9).

Studies have shown that the mortality rate of ASDH is up to 60%, and most patients with ASDH have no consciousness from the time of admission, and thus their prognosis is poor, regardless of whether or not surgery is performed10). The main dilemma is deciding whether or not to perform maximal care, or to refrain from aggressive treatment and perform palliative care instead11). Predicting the prognosis of patients diagnosed with ASDH is often difficult and leads to controversial surgical decisions. Various prognostic factors are implicated in the outcome of ASDH, including patient age, admission Glasgow Coma Scale (GCS), and associated injuries10,1214). In addition, radiological findings on the brain CT scans play an important role in management planning15).

The thickness of the ASDH is considered to be an important determinant of surgical decisions16,17). Recently, it has been reported that the outcome can be predicted by the ratio between hematoma thickness and MLS2,11). Although there are many models of prediction tools, there is no way to describe the prognosis objectively. We retrospectively investigated preoperative clinical status and imaging findings in ASDH patients who underwent hematoma evacuation in our institution. Based on this, we conducted the study to provide a communication tool for rationally reporting the patient’s prognosis to the caregivers.

MATERIALS AND METHODS

1. Study Design

A retrospective study was performed on 142 adult patients who received surgical decompression and hematoma evacuation after a diagnosis of traumatic ASDH at the Konkuk University Medical Center between August 2005 and March 2017. Data from all patients who met the study criteria were extracted from the database, including the patient’s sex and age, history, medications given, neurological examinations (including preoperative GCS and pupil abnormalities), thickness of the hematoma and MLS, presence of basal cistern obliteration, presence of subarachnoid hemorrhage (SAH) or contusional hemorrhage, and 1-month follow-up Glasgow Outcome Scale (GOS). Data were extracted, based on the inclusion criteria, from patients aged 18 years or older diagnosed with traumatic ASDH with a hematoma of thickness 5mm or more, who underwent unilateral craniotomy or craniectomy, among patients with a GCS of 8 or less at the time of admission. Patients younger than 18 years of age (n=1), having a GCS of 9 or greater at the time of admission (n=31), with a previous history of cranial surgery (n=7), with uncontrolled primary malignancy and a hematologic disorder (n=1) or with the Karnofsky Performance Score (KPS) score below 60 before the trauma (n=4) were excluded. Patients with a reason for choosing surgery for other hematomas with a mass effect rather than subdural hematoma (SDH) (n=5), spontaneous SDH (n=4), or patients with SDH due to other intracranial pathologies (n=2) were also excluded. Patients who underwent craniotomy on both sides or at the posterior fossa (n=9) were excluded, and patients with unavailable preoperative CT images (n=1) were also excluded. Patients who were lost to follow-up during treatment were also excluded (n=2). Finally, 75 patients were included in this study (Fig. 1). The age of the patients ranged from 22 to 83 years with a median age of 61 years (range, 22–83 years), and the male to female ratio was 53:22. The median preoperative GCS was 5. Falls (n=48) were the most common mechanism of ASDH, followed by pedestrian traffic accidents (n=13) and motorcycle traffic accidents (n=10). There were also some assault injuries (n=2).

Fig. 1

The flow of patient selection process using our inclusion criteria and exclusion criteria during the period from August 1, 2005 to March 31, 2017. SDH: subdural hematoma; ASDH: acute SDH; GCS: Glasgow Coma Scale; F-T-P: fronto-temporoparietal; KPS: Karnofsky Performance Score; PF: posterior fossa; ACA: anterior cerebral artery; f/u: follow up; pre OP: preoperative; CT: computed tomography.

The preoperative imaging findings were defined as follows. Measurements of the thickness of the hematoma and MLS were obtained at the level of the frontal horns. MLS was defined as the distance measured from the midline to the displaced septum pellucidum. The difference between the thickness of the hematoma and the MLS was expressed in millimeters. The state of the basal cistern was classified as follows: ‘patent’ if at least one side is open, ‘obliterated’ if neither is visible. Because the presence of SAH or contusional hemorrhage could be a predictive value for parenchymal damage, we checked for SAH or contusional hemorrhage on preoperative CT.

The surgical technique used to remove the ASDH was decided by the neurosurgeon during surgery. All patients who satisfied our inclusion criteria received decompressive craniectomy or craniotomy, which were determined based on their preoperative clinical status, CT findings, and operative findings. After surgical intervention, all patients were transferred to the intensive care unit for postoperative management and monitoring. Their intracranial pressure (ICP) was monitored in patients who had an ICP monitoring catheter. Only after extubation and vital functions were stabilized was the patient transferred to the ward for further management.

The GOS evaluates the patient’s functional outcome on a 5-point scale, from 1 (death) to 5 (good recovery). Patients were categorized in the poor functional outcome group if their 1-month follow-up GOS ranged from 1 (death) to 2 (vegetative state). One month after surgery, the patients were classified into the ‘Unfavorable’ group, which was measured by the GOS from 1 to 2, or the ‘Favorable’ group, which was measured by the GOS from 3 to 5.

2. Statistical Analysis

Associations between clinic-radiologic variables and prognosis were evaluated with the χ2 test. A multivariate logistic regression model was used to adjust for covariates which showed p<0.2 on the univariate tests. Variables with p<0.05 were defined as significantly relevant. These analyses were performed using the Statistical Package for the Social Sciences version 17.0 (SPSS Inc., Chicago, IL, USA). For the variables showing significance in multivariate analysis, recursive partitioning analysis (RPA) was done with free software (R version 3.3.3; The R Foundation for Statistical Computing, Vienna, Austria) to create the recursive decision tree with the split criteria of p<0.05. The final nodes were grouped according to their 1-month mortality and favorable/unfavorable prognosis rate.

RESULTS

1. Prognostic Factors in Severe Traumatic Acute SDH

One month after surgery, 54 patients (72%) had poor GOS 1 to 2 (unfavorable outcome) and 21 patients (28%) had GOS 3 to 5 (favorable outcome). Patients with a low preoperative GCS (3–5) showed a statistically high incidence of an unfavorable outcome with a low GOS (1–2) a month after surgery (p<0.001). Pupillary abnormalities (bilateral fixed dilated pupils) on admission was associated with unfavorable outcomes (p= 0.006). A statistically significant unfavorable outcome was also observed in patients with loss of self-respiration during admission (p=0.001). The time from the onset of the injury to the start of surgery varied from a minimum of 1.2 hr to a maximum of 46 hr, and an unfavorable prognosis was observed in patients who underwent surgery within 6 hr (p=0.042).

The imaging findings that were associated with the prognosis of the patients were also observed in this study. An unfavorable outcome could be predicted if the thickness of the hematoma was 13mm or more (p=0.025), there was basal cistern obliteration (p<0.001), or SAH or contusional hemorrhage (p=0.030) was observed on preoperative CT scan images.

A multivariate analysis was performed with significant univariate predictors entered into multivariate regression models. Lower admission GCS (p=0.030; 95% confidence interval[CI]= 1.24–68.33), presence of SAH or contusional hemorrhage (p= 0.050; 95% CI=1.00–56.97), observed basal cistern obliteration (p=0.018; 95% CI=1.40–37.19), and a greater thickness of hematoma (p=0.037; 95% CI=1.12–35.66) were associated with unfavorable outcomes of ASDH patients. The observed data during the hospital period are shown in Table 1.

Variables related to outcome in patients who underwent surgery for traumatic acute subdural hematoma

2. Decision Tree for the Prognosis of Acute SDH

Using the above-mentioned significant prognostic factors (GCS 3–5, basal cistern obliteration, hematoma thickness ≥13mm, presence of SAH or contusional hemorrhage) for ASDH, a recursive decision tree comprising 75 patients was created. Figure 2 shows 3 terminal nodes based on the preoperative KPS scores and the status of the basal cistern. Based on the terminal nodes, we categorized them into three groups. Independent of the radiologic findings, patients with a GCS of 3 to 5 can expect the worst outcomes (Group A). Among the patients with a GCS of 6 to 8, the patients without obliteration of the basal cistern (Group C) had a better 1 month clinical prognosis than those with obliteration of the basal cistern (Group B). The 1-month favorable outcome rates for each group were 4.9% in group A, 23.1% in group B, and 76.2% in group C. The same decision tree was drawn when applying the 1-month mortality rates. The 1-month mortality rates in groups A, B, and C were 68.3%, 53.8%, and 0%, respectively.

Fig. 2

The decision tree which is constructed by applying the significant factors in the ultivariate analysis to the recursive partitioning model. GCS: Glasgow Coma Scale. GOS: Glasgow Outcome Score.

3. Illustrative Cases

The patient had a relatively high GCS of 8 and basal cistern obliteration was observed (Fig. 3A). In this case, the patient was categorized as Group B and we predicted an unfavorable outcome. As predicted in the decision tree, this patient died 19 days after admission. The second patient had a low GCS of 5 on admission, and a basal cistern obliteration was observed in the CT scan image (Fig. 3B). According to the decision tree, the patient was in Group A and an unfavorable outcome was predicted. However, the patient had a favorable outcome with a GOS of 4 months after the operation. The third patient had a low GCS of 3 and was classified as Group A, suggesting an unfavorable outcome. However, a month after surgery, his GOS of 3 was checked and the patient had a favorable outcome. Basal cistern obliteration was not observed on a CT scan of this patient (Fig. 3C). Therefore, it is important to be aware that an exceptional situation may arise when determining the required type of surgery based on this diagram or when explaining it to the patient’s caregiver.

Fig. 3

Preoperative (Pre OP) and postoperative (Post OP) computed tomography (CT) of acute subdural hematoma patients who underwent surgery. (A) Axial CT images of patient with preoperative Glasgow Coma Scale (GCS) 8 and basal cistern obliteration. (B) Axial CT images of patient with preoperative GCS 5 and basal cistern obliteration. (C) Axial CT images of patient with preoperative GCS 5 and patent basal cistern.

DISCUSSION

The RPA classification not only provides simple and intuitive information on risk grouping, but also allows grouping of patients with the same risk and allows the clinician to predict the clinical prognosis of a particular patient population. However, it is clear that the ability to identify prognostic factors is low when the variables are large compared with the multivariate regression model18). Thus, we first identified four prognostic factors by analyzing prognostic factors using chi-square tests and multiple logistic regression models and applied them to the RPA model.

Preoperative GCS was the most powerful splitting criteria in our model. It is well-known that GCS not only directly reflects brain damage and clinical status, but also provides information about survival during follow-up10,19,20). In previous studies, preoperative GCS has been reported to be associated with the outcome of patients with TBI (Table 2)10,14,17,2023). A low GCS on admission is known to be an important prognostic factor10,23). It is known that mortality is high in patients with GCS 3 to 5 at admission22). Previously published data have shown mortality rates of 60% to 84% in patients with a preoperative GCS of 3 to 512,13). Because the results were statistically significant, the preoperative GCS was used as a factor in the derivation of the RPA model and as a result it was found to be an important axis in our RPA decision tree. In our RPA decision tree, patients with preoperative GCS 3 to 5 were assigned to group A, and 95.1% of the 41 patients in group A showed an unfavorable outcome indicating GOS 1 to 2 a month after surgery.

Prognostic factors of acute subdural hematoma known from previous literature

Our RPA model provides two nodes separated by the status of the basal cistern in patients with a GCS of 6 to 8. Studies have shown that the presence or absence of basal cistern obliteration on the CT scan image is strongly associated with the prognosis of the patient14,16,24). Toutant et al.21) reported that a 77% mortality rate was observed when basal cistern obliteration existed. In our study, 90.2% (37/41) of patients showed unfavorable outcomes with basal cistern obliteration, consistent with previous findings. Among the patients with GCS 6 to 8, patients with basal cistern obliteration were placed in group B and those without basal cistern obliteration were placed in group C. According to our RPA tree, the outcomes of these groups are clearly different. In Group B, 76.9% showed unfavorable outcomes with GOS 1 to 2 at 1 month postoperatively and 53.8% mortality. In group C, 76.2% showed a favorable outcome with GOS 3 to 5 and 0% mortality.

Among the preoperative imaging findings, results that could be regarded as prognostic factors were also derived. It has been reported that unfavorable outcomes are observed when SAH is present on CT scan images before surgery16,21,22). We noted the outcomes of patients with SAH or contusional hemorrhage on preoperative CT scans. A total of 79% of these patients showed unfavorable outcomes. We also examined data from our patients based on a previous study indicating that a greater thickness of hematoma could be used as a poor prognostic factor16,17). A mean hematoma thickness of 13mm was detected in patients satisfying our inclusion criteria and an unfavorable outcome was observed when the thickness of hematoma was greater than 13mm. However, despite the significant statistical results of the two factors mentioned above, the two factors failed to be a splitting criterion in RPA analysis because they did not have priority in executing the RPA.

In our study, an unfavorable outcome was observed in patients who underwent surgery within 6 hr of the injury (p=0.042), although a previous study showed a favorable outcome in patients who underwent surgery within 4 hr of admission14). This is probably due to the severe brain damage in patients who underwent surgery within 6 hours because of ASDH, contusional hemorrhage, and DAI. In addition, preoperative basal cistern obliteration or greater thickness of a hematoma was associated with a poor prognosis, which was associated with severe clinical symptoms, suggesting a poor prognosis.

Previous studies have reported that MLS and MLS/thickness of hematoma ratio are strongly associated with prognostic factors2,11,14,16,17), but our study did not show any significant association. We compared the results of other studies with MLS or MLS/thickness of hematoma ratio as prognostic predictive factors. In our study, bilateral SAH or contusional hemorrhage were observed in most patients with severe brain injury. Because mass effect was observed on both sides, we presumed that MLS could be compensated by bilateral mass effect. In addition, the preoperative GCS of the patients in our study was lower (patients with a GCS of 3 to 5 accounted for 55% of all patients) than in other study groups, suggesting that the severity of brain damage was high. Therefore, it could be assumed that the influence of MLS or MLS/thickness of hematoma ratio is not observed in our study more often than in other studies. However, since the number of patients in our study is relatively small, further investigation is mandatory to reveal the effect of MLS/thickness of hematoma ratio to the outcome of ASDH patients. Some studies have shown that antithrombotics21) and serum albumin levels23) are associated with patient prognosis. However, there was no difference in outcomes according to the serum albumin level or the use of antiplatelet/anticoagulation medication in this study.

There are some limitations in our study. First, we applied strict inclusion criteria and therefore selection bias may have occurred. However, since ASDH is often accompanied by various injury mechanisms and involves large spectrum of heterogeneous disease entities. For the sake of homogenous study group, we excluded 47.2% of patients showing relatively good neurological status or unusual clinical status. Second, as shown in the illustrative cases, some cases do not match the diagram we have proposed. When performing CT scans in the emergency room, we did not perform thin-section CT scans such as CT angiograms in all patients. In most cases, a CT scan was done in a setting with an interval of 5mm per section. Thus, even if basal cistern is observed, there is a possibility of over-estimation as if obliteration existed. Moreover, because the concept of basal cistern obliteration itself is vague, evaluating the degree of basal cistern obliteration is not easy. One of the two patients with a favorable outcome who belonged to Group A had no basal cistern obliteration observed on the preoperative CT scan. Third, the number of patients studied is small. Because of this, variables that were meaningful in multivariate analysis were not selected and failed to form a decision tree. Therefore, for further study of ASDH, it is necessary to increase the number of patients and perform more thorough imaging studies.

CONCLUSION

There have been studies on the prognostic predictors of ASDH and some prognostic factors have been identified in this study. We were able to establish a model for predicting prognosis in patients with ASDH through preoperative GCS and basal cistern obliteration on CT scan images. It is expected that this model will not only provide objective information when we make decisions about treatment, but it can also be a useful tool when discussing the patient’s prognosis with the patient’s caregivers.

Notes

CONFLICTS OF INTEREST

No potential conflict of interest relevant to this article was reported.

References

1. Alagoz F, Yildirim AE, Sahinoglu M, Korkmaz M, Secer M, Celik H, et al. Traumatic acute subdural hematomas: analysis of outcomes and predictive factors at a single center. Turk Neurosurg 27:187–191. 2017;
2. Bartels RH, Meijer FJ, van der Hoeven H, Edwards M, Prokop M. Midline shift in relation to thickness of traumatic acute subdural hematoma predicts mortality. BMC Neurol 15:220. 2015;
3. Bullock MR, Chesnut R, Ghajar J, Gordon D, Hartl R, Newell DW, et al. Surgical management of acute subdural hematomas. Neurosurgery 58:S16–S24. 2006;
4. Cook EF, Goldman L. Empiric comparison of multivariate analytic techniques: advantages and disadvantages of recursive partitioning analysis. J Chronic Dis 37:721–731. 1984;
5. Dent DL, Croce MA, Menke PG, Young BH, Hinson MS, Kudsk KA, et al. Prognostic factors after acute subdural hematoma. J Trauma 39:36–42. discussion 42–33. 1995;
6. Han MH, Ryu JI, Kim CH, Kim JM, Cheong JH, Yi HJ. Radiologic findings and patient factors associated with 30-day mortality after surgical evacuation of subdural hematoma in patients less than 65 years old. J Korean Neurosurg Soc 60:239–249. 2017;
7. Hatashita S, Koga N, Hosaka Y, Takagi S. Acute subdural hematoma: severity of injury, surgical intervention, and mortality. Neurol Med Chir (Tokyo) 33:13–18. 1993;
8. Kalanithi P, Schubert RD, Lad SP, Harris OA, Boakye M. Hospital costs, incidence, and inhospital mortality rates of traumatic subdural hematoma in the United States. J Neurosurg 115:1013–1018. 2011;
9. Karibe H, Hayashi T, Hirano T, Kameyama M, Nakagawa A, Tominaga T. Surgical management of traumatic acute subdural hematoma in adults: a review. Neurol Med Chir (Tokyo) 54:887–894. 2014;
10. Kim JJ, Gean AD. Imaging for the diagnosis and management of traumatic brain injury. Neurotherapeutics 8:39–53. 2011;
11. Kim KH. Predictors for functional recovery and mortality of surgically treated traumatic acute subdural hematomas in 256 patients. J Korean Neurosurg Soc 45:143–150. 2009;
12. Kolias AG, Scotton WJ, Belli A, King AT, Brennan PM, Bulters DO, et al. Surgical management of acute subdural haematomas: current practice patterns in the United Kingdom and the Republic of Ireland. Br J Neurosurg 27:330–333. 2013;
13. Leitgeb J, Mauritz W, Brazinova A, Janciak I, Majdan M, Wilbacher I, et al. Outcome after severe brain trauma due to acute subdural hematoma. J Neurosurg 117:324–333. 2012;
14. Marmarou A, Lu J, Butcher I, McHugh GS, Murray GD, Steyerberg EW, et al. Prognostic value of the Glasgow Coma Scale and pupil reactivity in traumatic brain injury assessed pre-hospital and on enrollment: an IMPACT analysis. J Neurotrauma 24:270–280. 2007;
15. Moussa WMM, Khedr WM, Elwany AH. Prognostic significance of hematoma thickness to midline shift ratio in patients with acute intracranial subdural hematoma: a retrospective study. Neurosurg Rev epub ahead of print. 2017;10.1007/s10143-017-0873-5.
16. Ono J, Yamaura A, Kubota M, Okimura Y, Isobe K. Outcome prediction in severe head injury: analyses of clinical prognostic factors. J Clin Neurosci 8:120–123. 2001;
17. Paci GM, Sise MJ, Sise CB, Sack DI, Swanson SM, Holbrook TL, et al. The need for immediate computed tomography scan after emergency craniotomy for head injury. J Trauma 64:326–333. discussion 333–324. 2008;
18. Rasouli MR, Rahimi-Movaghar V. Time-to-treatment and mortality in patients with acute subdural hematoma. Ann Surg 257:e8. 2013;
19. Servadei F, Nasi MT, Giuliani G, Cremonini AM, Cenni P, Zappi D, et al. CT prognostic factors in acute subdural haematomas: the value of the ‘worst’ CT scan. Br J Neurosurg 14:110–116. 2000;
20. Taussky P, Hidalgo ET, Landolt H, Fandino J. Age and salvage-ability: analysis of outcome of patients older than 65 years undergoing craniotomy for acute traumatic subdural hematoma. World Neurosurg 78:306–311. 2012;
21. Toutant SM, Klauber MR, Marshall LF, Toole BM, Bowers SA, Seelig JM, et al. Absent or compressed basal cisterns on first CT scan: ominous predictors of outcome in severe head injury. J Neurosurg 61:691–694. 1984;
22. Tsang KK, Whitfield PC. Traumatic brain injury: review of current management strategies. Br J Oral Maxillofac Surg 50:298–308. 2012;
23. Wilberger JE Jr, Harris M, Diamond DL. Acute subdural hematoma: morbidity, mortality, and operative timing. J Neurosurg 74:212–218. 1991;
24. Yu P, Tian Q, Wen X, Zhang Z, Jiang R. Analysis of long-term prognosis and prognostic predictors in severe brain injury patients undergoing decompressive craniectomy and standard care. J Craniofac Surg 26:e635–e641. 2015;

Article information Continued

Fig. 1

The flow of patient selection process using our inclusion criteria and exclusion criteria during the period from August 1, 2005 to March 31, 2017. SDH: subdural hematoma; ASDH: acute SDH; GCS: Glasgow Coma Scale; F-T-P: fronto-temporoparietal; KPS: Karnofsky Performance Score; PF: posterior fossa; ACA: anterior cerebral artery; f/u: follow up; pre OP: preoperative; CT: computed tomography.

Fig. 2

The decision tree which is constructed by applying the significant factors in the ultivariate analysis to the recursive partitioning model. GCS: Glasgow Coma Scale. GOS: Glasgow Outcome Score.

Fig. 3

Preoperative (Pre OP) and postoperative (Post OP) computed tomography (CT) of acute subdural hematoma patients who underwent surgery. (A) Axial CT images of patient with preoperative Glasgow Coma Scale (GCS) 8 and basal cistern obliteration. (B) Axial CT images of patient with preoperative GCS 5 and basal cistern obliteration. (C) Axial CT images of patient with preoperative GCS 5 and patent basal cistern.

Table 1

Variables related to outcome in patients who underwent surgery for traumatic acute subdural hematoma

No. of unfavorable outcome No. of patients Univariate Multivariate

p-value HR 95% CI
Sex 0.928
 Male 38 53
 Female 16 22
Age 0.399
 <60 years 25 37
 ≥60 years 29 38
Comorbidities 0.343
 No 35 51
 Yes 19 24
Antiplatelet or anticoagulation 0.680
 No 50 70
 Yes 4 5
Accompanying injuries 0.544
 No 43 61
 Yes 11 14
Lucid interval 0.958
 No 10 14
 Yes 44 61
GCS <0.001 0.030 9.203 1.239–68.333
 3–5 15 34
 6–8 39 41
Pupil change 0.006 0.911 0.908 0.169–4.884
 None or ipsilateral 9 19
 Bilateral 45 56
Loss of self respiration 0.001 0.512 2.460 0.167–36.271
 No 30 50
 Yes 24 25
Time interval from onset of complaint to surgery 0.042 0.431 2.151 0.319–14.491
 >6 hours 7 14
 ≤6 hours 47 61
Spot sign in CT angiography 0.891
 No 51 71
 Yes 3 4
SAH or contusional hemorrhage 0.030 0.050 7.557 1.003–56.968
 No 10 19
 Yes 44 56
Basal cistern obliteration <0.001 0.018 7.215 1.400–37.188
 No 17 34
 Yes 37 41
Thickness of hematoma 0.025 0.037 6.314 1.118–35.659
 <13 mm 23 38
 ≥13 mm 31 37
Midline shift 0.536
 <14 mm 24 35
 ≥14 mm 30 40
Midline shift/Thickness of hematoma ratio 0.445
 <1 31 41
 ≥1 23 34
Skull fracture 0.440
 No 18 27
 Yes 36 48
Serum albumin level 0.665
 <3.5 g/dL 6 7
 ≥3.5 g/dL 48 68

HR: hazard ratio; CI: confidence intervals; GCS: Glasgow Coma Scale; CT: computed tomography; SAH: subarachnoid hemorrhage.

Table 2

Prognostic factors of acute subdural hematoma known from previous literature

References No. of patients Overall mortality Proposed prognostic factor
Toutant et al., 198421) 218 77.0% Basal cistern absence
Dent et al., 19955) 211 52.0% Admission GCS, midline shift, basal cistern absence, hematoma evacuated within 4 hours
Servadei et al., 200019) 206 46.0% Thickness of hematoma, midline shift, basal cistern absence, presence of SAH
Ono et al., 200116) 272 N/A Admission GCS, pupillary abnormalities, presence of SAH
Kim, 200911) 256 39.8% Mechanism of injury, admission GCS, pupillary abnormalities, thickness of hematoma, volume of hematoma, midline shift
Leitgeb et al., 201213) 863 47.6% Age, admission GCS, pupillary abnormalities
Bartels et al., 20152) 59 50.8% Midline shift in relation to thickness of hematoma
Yu et al., 201524) 1,780 N/A Admission GCS, serum albumin level
Alagoz et al., 20171) 99 64.1% Admission GCS, thickness of hematoma
Han et al., 20176) 318 22.0% Admission GCS, use of antithrombotics, history of diabetes mellitus, presence of SAH
Moussa et al., 201715) 67 23.9% Midline shift/Thickness of hematoma ratio

GCS: Glasgow Coma Scale; SAH: subarachnoid hemorrhage.