TBI is defined as head trauma associated with loss of consciousness for more than 24 h, intracranial hemorrhage, or depressed skull fracture. Approximately 5–7 % of all patients with TBI will experience early seizures that are defined as seizures occurring within 1 week following the trauma [7–9]. The incidence of early seizures in patients with acute traumatic SDH is 11–30 % when prophylactic AEDs are not used [7, 9–11], with 60–70 %of those occurring in the first 24 h following the insult [8, 9, 11]. Furthermore, acute SDH is present in about 60 % of all patients with TBI experiencing early seizures, while 50 % of these patients suffer from brain contusion and 10 % from an epidural hematoma [7]. Patients with TBI and an acute SDH are approximately three times more likely to develop early seizures than TBI patients without an SDH [7]. In addition to being at an increased risk of developing early seizures, patients with acute SDH carry an increased risk of developing late posttraumatic seizures, defined as seizures occurring more than 1 week following the TBI. Patients with acute SDH that is large enough to require surgical intervention have a 44 % risk of developing seizures within 2 years following the TBI. This risk decreases to approximately 20 % in patients with an acute SDH that did not require surgical intervention [10]. The latter rate is similar to the one experienced by TBI without SDH. The risk of developing seizures in patients with chronic SDH ranges between 2.3 and 17 % depending on the type and severity of the head injury [12–16]. Clinical and radiological risk factors for developing seizures associated with chronic SDH include brain atrophy, mixed-density SDH, prior stroke, and low Glasgow Coma Scale (GCS) at presentation [16–18]. In patients with chronic SDH who underwent surgical intervention, the incidence of seizures is between 5 and 22 % in the acute postoperative period [17–21]. Predictors of seizures in this group include low GCS on admission, low postoperative GCS, acute-on-chronic SDH, and open craniotomy [17, 18]. The degree of midline brain shift and the volume of the SDH are reported not to predict the occurrence of seizures [17]. Seizures in this patient population had an impact on the short-term, postoperative outcome but did not impact the long-term outcome [17] .

SDH is commonly seen along the cerebral convexities, the falx cerebri, and the tentorium cerebelli. It undergoes a typical temporal evolution on both computed tomography (CT) and magnetic resonance (MR) imaging [22]. An acute SDH appears as a hyperintense crescent-shaped collection on a head CT scan (Fig. 5.1a). On rare occasions, such as in patients with anemia, disseminated intravascular coagulopathy, or with tears in the arachnoid membrane causing dilution from the cerebrospinal fluid, an acute SDH may be isodense or hypodense on a head CT scan. Subacute SDH appears isodense to gray matter, making it challenging to recognize on a head CT scan, especially if there are bilateral SDHs (Fig. 5.1b). Chronic SDH appears homogenously hypodense on head CT scan (Fig. 5.1c). Acute-on-chronic SDH has a bilayered appearance on head CT with a hypodense layer in the less position-dependent portion and hyperdense component in the position-dependent portion (Fig. 5.1d). Magnetic resonance imaging (MRI) is more sensitive than CT imaging for the detection of a small SDH. Acute SDH is isointense on T1-weighted images (WI) and hypointense on T2-WI. Subacute SDH is hyperintense on T1-WI and T2-WI (Fig. 5.2). Over time, the hyperintensity diminishes, and chronic SDH appears as isointense or hypointense on T1-WI and hypointense on T2-WI.

SDH is formed through the extravasation of blood within the dura–arachnoid interface layer that splits the space while leaving a few tiers of dural border cells over the arachnoid. These cells that cover the internal surface of the hematoma may proliferate later to form a neo-membrane. The outer membrane may expand due to repeated hemorrhage and in conjunction with the inner membrane leading to a capsule formation (Fig. 5.3). Acute SDH is usually absorbed within a few weeks but might persist to become subacute SDH and evolve into chronic SDH.

There are several proposed mechanisms to explain the pathophysiology of seizures in patients with SDH. The traditional proposed mechanism for epileptogenicity in these patients is direct cortical irritation and hyperexcitability [23]. A mixed-density SDH on head CT scan is due to the presence of fresh erythrocytes. This may lead to increased fibrinogen degradation products in patients with acute on top of chronic SDH, which may influence the permeability of the SDH capsule membrane affecting the brain parenchyma and causing seizures [18]. Also, the volume of a mixed-density SDH tends to be larger than a low-density SDH, which may potentially generate more mass effect and pressure on the brain parenchyma resulting in seizures [24]. Another postulated mechanism is that the SDH may reduce the regional cerebral blood flow resulting in cortical ischemia and hyperexcitability [24]. The surgical techniques employed in the treatment of SDH might also be important in the pathogenesis of seizures. Although the SDH capsule is thought to be the source of epileptogenesis , the incidence of postoperative seizures in patients who underwent capsulotomy was higher than in those who underwent burr-hole drainage [13, 17, 20]. This suggests that cortical injury and gliosis due to either surgical intervention or the SDH itself can result in late seizures [17]. Another important factor is that SDH is often associated with other brain injuries such as cerebral contusions, subarachnoid hemorrhage, cerebral ischemia, and increased intracranial pressure with midline shift. All these coexisting brain injuries have been independently associated with seizures as well. Considering all the possible mechanisms by which seizures can develop in patients with SDH, health-care providers should have a very low threshold for considering seizure diagnosis in patients with SDH based on the clinical presentation and course.

There are limited data detailing the clinical seizure types encountered in patients with SDH. Considering the focal nature of the lesion, the commonly expected seizure types are various forms of partial-onset seizures with or without secondary generalization. However, the available literature regarding seizure types in patients with acute SDH indicates that about 60–80 % of the seizures following subdural hematoma are generalized, while only 20–40 % are focal seizures [16, 18]. These reports raise questions regarding the certainty in seizure classification between primary generalized and secondary generalized convulsive seizures based purely on the available clinical seizure semiology. In addition to single seizure, patients with SDH are at risk of convulsive and nonconvulsive status epilepticus. Although all forms of status epilepticus are not uncommonly encountered in neuro-intensive care units, specific medical literature addressing status epilepticus in patients with SDH is lacking. Status epilepticus (convulsive or nonconvulsive) is reported in 2–8 % of patients with severe TBI [25–30]. However, the real incidence is likely higher, perhaps in the range of 20 %, due to the fact that about 50 % of seizures may be subtle and clinically undetectable [25]. This makes electroencephalography (EEG), particularly continuous EEG monitoring, essential for detecting and managing clinical and subclinical seizures in this population.

The EEG is a necessary tool in evaluating patients with SDH as it can provide both diagnostic and prognostic information. It often shows epileptiform and non-epileptiform findings, which can help in lateralizing and localizing the site of hemorrhage. The reported non-epileptiform EEG findings include generalized or focal slowing patterns, focal voltage attenuation and asymmetries, and intermittent rhythmic delta activity (IRDA). These findings are nonspecific and reflect the lateralization and localization of the SDH and its structural effects. The diagnostic role of EEG in SDH has diminished greatly with the advent of CT technology in the mid-1970s. However, it remains of utmost importance in the evaluation and management of clinical as well as subclinical seizures, and it can sometimes provide prognostic information in patients with SDH. Very few studies report the EEG findings in patients with SDH without confounding cerebral injuries. Rudzinski et al. [31] reported EEG findings in patients with acute SDH as an isolated lesion. Patients in this study underwent surgical evacuation of acute SDH and postoperatively either had clinical seizures or other neurological deterioration requiring an EEG recording. Epileptiform discharges were common, occurred in 87 % of the cohort, but were not necessarily lateralizing to the side of SDH. Bilateral epileptiform discharges were noted in 66 % of these patients [31]. In the absence of cerebral contusions or prior brain disease, epileptiform discharges contralateral to the SDH could be related to disturbed function of the contralateral hemisphere from mass effect imposed by the SDH [32]. Additionally, contralateral epileptiform discharges may also be markers of severe diffuse cerebral dysfunction in these patients (Fig. 5.4). The presence of midline epileptiform discharges on the EEG correlated significantly with the presence of midline shift on brain imaging studies. This was proposed to be secondary to vascular insufficiency involving the watershed regions of the medial cerebral hemisphere [31]. In addition to focal epileptiform discharges, 43 % of these patients had periodic lateralized epileptiform discharges (PLEDs) on routine EEG (Fig. 5.5). PLEDs tended to be much more localizing to the side of the SDH than focal epileptiform discharges and always correlated to an underlying hematoma. Patients with SDH and PLEDs on the EEG are also more likely to have a midline shift on brain imaging studies [33, 34]. Focal epileptiform discharges and PLEDs were more likely to be noted in patients with acute SDH as compared to patients with acute on top of chronic SDH. Focal or multifocal subclinical electrographic seizures were also noted in 12 % of these patients undergoing EEG postoperatively (Fig. 5.6) [31]. Bihemispheric diffuse slowing is the most common background abnormality present on the EEG of 92 % of patients with SDH. It is often asymmetrical and more prominent on the hemisphere that is ipsilateral to the hematoma. Focal slowing is most likely to be noted in the temporal regions and is thought to be secondary to mass effect compromising the regional cerebral blood flow [35]. Amplitude asymmetries are also a common finding on the EEGs of patients with SDH. Decreased amplitude is expected on the side ipsilateral to the SDH prior to evacuation, but it might persist for a prolonged period following evacuation (Fig. 5.4). This is postulated to be secondary to compression of the ipsilateral subcortical tracts [32]. Increased amplitude of the EEG activity ipsilateral to the SDH may be due to a breach rhythm which commonly occurs following surgical burr holes and craniotomies (Fig. 5.7). In addition to background slowing and amplitude asymmetries, frontal intermittent rhythmic delta activity (FIRDA) and bihemispheric frontally dominant triphasic slow waves might be observed on the EEG of patients with SDH. FIRDA and triphasic slow waves are more likely to be encountered in patients with acute on top of chronic SDH than in patients with acute SDH. Focal IRDAs are of special interest if encountered as they correlate with focal potential epileptogenicity (Fig. 5.8).

The mortality rate of patients with acute SDH ranged from 40 to 60 % [36] in the early 1990s and more recently decreased to about 20 % [37]. The mortality rate of chronic SDH is between 3 and 10 % [38, 39]. Predictors of increased risk of mortality include older age, low admission GCS, prolonged hospitalization, coagulopathy, mechanical ventilation, and multiple comorbidities. Acute SDH can recur in 14–22 % of cases [40, 41]. This is more likely to happen if there is an active extravasation of contrast on head CT angiography (CTA) [42]. Chronic SDH may recur in 6–24 % [38, 43]. Recurrence is more likely to happen in patients with preoperative hematoma volume greater than 115 mL and postoperative residual hematoma volume greater than 80 mL, postoperative midline shift greater than 5 mm, preoperative seizures, and in patients on anticoagulant therapy [38, 44–46]. In addition, patients with diabetes mellitus, multilocular hematoma, and massive postoperative subdural air are at an increased risk of SDH recurrence. In recent years, the in-hospital mortality from SDH has decreased nationwide, but there is a major increase in discharges to hospice and long-term care facilities leading to increased consumption of health-care resources [47].