Patient Awareness of Complex Partial Seizures


Address correspondence and reprint requests to Dr. Kyoung Heo, Department of Neurology, Yonsei University College of Medicine, Severance Hospital, Shinchon-dong 134, Seodaemoon-ku, Seoul 120-752, Korea. E-mail:


Summary: Purpose: To assess self-awareness of complex partial seizures (CPSs) in unselected epilepsy patients through a thorough interview.

Methods: The study comprised 134 patients at our epilepsy clinic, whose CPSs had been documented by the patient's family members. We investigated the proportion and characteristics of patients unaware of their CPSs compared to those who were, and we monitored the evolution of unawareness of CPSs during the follow-up.

Results: Thirty-one (23%) patients were assigned to the unawareness group (complete, 23; incomplete, 8) and 103 (67%) patients to the awareness group. Patients in the unawareness group were older and had a later age of onset than patients in the awareness group. Interictal epileptiform discharges (IEDs) localized primarily to the temporal region and were more frequently detected in the unawareness group (94%) than the awareness group (55%). Bilateral independent IEDs were found more frequently in the unawareness group than in the awareness group (48% vs. 13%). The bilateral presence of lesions was also more frequent in the unawareness group than the awareness group (16.1% vs. 4.9%). Six (26%) of 23 patients with complete unawareness of their CPSs had experienced awareness of CPSs during the follow-up. Two of these patients even experienced the emergence of de novo aura.

Conclusions: Our results indicate that a significant number of epilepsy patients are not aware of their CPSs. Unawareness of CPSs may be related to bitemporal dysfunction and a rapid and complete loss of consciousness caused by rapid spread of ictal discharges to the contralateral hemisphere in association with bilateral independent IEDs and bilateral presence of lesions.

The management of patients with epilepsy relies heavily on self-reported seizure frequency. Accurate self-reporting requires the patient's awareness of seizures. Patient unawareness of their seizures might result in an underreporting of seizure frequency that can compromise therapeutic trials.

Prior studies have reported that a significant proportion of patients were either completely or partially unaware of their seizures (Gotman, 1990; Blum et al., 1996; Tatum et al., 2001). However, because all these studies used continuous EEG to monitor seizures and focused primarily on patients with refractory epilepsies, the relevance of these studies to clinical practice is not clear. Therefore, we conducted a prospective study in which we performed thorough interviews of patients visiting our epilepsy clinic to investigate the incidence and clinical variables affecting the self-perception of seizures.


We prospectively investigated 166 consecutive patients, who had experienced complex partial seizures (CPSs) and visited the outpatient epilepsy clinic of Severance Hospital by a physician (K. Heo) from April 2000 to October 2004. CPS was defined as partial onset seizure with impaired consciousness or lack of responses but without overt convulsive movements. Patients that had CPSs only during sleep (n = 11), overt mental retardation or impaired communication skills (n = 10), or a total number of CPSs ≤3 while awake (n = 11) were excluded because these patients were unlikely to give reliable information regarding their perception of CPSs. To evaluate each patient's awareness of CPSs, at their initial visit patients underwent a thorough interview that included analysis with epilepsy-dedicated, high-resolution magnetic resonance imaging (MRI) and EEG with nasopharyngeal electrodes. This study was approved by Institutional Review Board of Severance Hospital.

Three questions were administered to evaluate the patient's awareness of CPSs which had been documented by the family members. The patient was asked, (1) “Even if your family members did not tell you the occurrence of events (CPSs), have you been aware of amnesic episodes (CPSs) which occurred recently?” The family members were asked, (2) “Do you think that the patient is aware of having seizures if you did not tell the occurrence of the events to the patient?” and (3) “Has the patient reported nothing wrong although you told the patient that the events had occurred?” If the patients answered “No” for Question 1, and their family members did “No” for Question 2, and “Yes” for Question 3, they were classified into the complete unawareness group (never aware of CPSs). If the patients answered “Yes” for Question 1, and their family members did “Yes” for Question 2 and “No” for Question 3, they were classified into the awareness group (always aware of CPSs). In cases of being sometimes aware of CPSs, they were classified into the incomplete unawareness group.

The analysis of interictal epileptiform discharges (IEDs) was based on the first scalp EEGs that were performed at our outpatient epilepsy clinic. Bilateral IEDs were defined if at least one interictal discharge appeared independently over each side. Lobar localization of epileptogenic foci was based on ictal semiology, EEG, and MRI findings (Commission on Classification and Terminology of the International League Against Epilepsy, 1989; Lee et al., 2002). Lateralization of epileptogenic foci was based on ictal semiology suggesting strong lateralizing semiological features, such as lateralized somatosensory, visual, and convulsive seizures, forced head-version before secondary generalization, unilateral ictal dystonia, postictal dysphasia, ictal speech, unilateral blinking, and ictal vomiting (Rosenow and Luders, 2001), lateralized EEG findings (>80% predominance of IEDs or consistent unilateral slowing), or unilateral MRI abnormalities.

Involvement of mesial temporal structures was investigated by visual inspection in patients with visible focal lesions on MRI. Additionally, during the subsequent follow-up we investigated the evolution of unawareness in patients with complete unawareness at their initial visit, when their recurrence of CPSs had been documented by their family members.

We used Student's t-test to analyze continuous variables and the χ2 test or Fisher's exact test to examine categorical variables. The significance level was set at p < 0.05. To account for the joint effects of several variables, a logistic regression model was performed.


A total of 134 patients (age ranged from 13 to 72 years) were investigated for the awareness of CPSs. Twenty-three (17.2%) patients were assigned to the complete unawareness group, eight (6.0%) patients to the incomplete unawareness group, and 103 (76.9%) patients to the awareness group. Because the number of patients in the incomplete unawareness group was small, we combined this group with the unawareness group in subsequent analysis of variables related to seizure unawareness. Fifty-three (39.6%) patients had not taken antiepileptic drugs (AEDs) at their initial visit to the clinic; patients with complete unawareness had not been treated with AEDs more frequently compared with those with awareness (65.2% vs. 35.9%, p = 0.01). Age at onset of epilepsy (p = 0.007) and age at the initial visit (p = 0.005) was greater in the unawareness group than in the awareness group. Experiencing aura is the hallmark of self-perception of seizures, and consistent with this no patients in the complete unawareness group reported experiencing aura. Other demographic characteristics, such as seizure duration, sex, and seizure type were similar in both the unawareness and awareness groups (Table 1).

Table 1. Demographic data
 Unawareness groupAwareness groupTotal
  1. ap= 0.005; bp= 0.007.

  2. CPS, complex partial seizure.

Female (%)14 (45.2)46 (44.7)60 (44.8)
Age at visit (yr ± SD)a42.5 ± 14.833.6 ± 15.335.7 ± 15.6
Age at onset (yr ± SD)b32.0 ± 18.422.9 ± 15.525.0 ± 16.6
Duration of epilepsy (month ± SD)123.6 ± 131.8129.3 ± 111.9128.0 ± 116.3
Seizure type (CPS only) (%)19 (61.3)54 (52.4)73 (54.5)
Aura experience (%) 5 (16.1)81 (78.6)87 (64.9)
Untreated patients (%)16 (48.4)37 (35.9)53 (39.6)

EEG and MRI findings, and epilepsy classification

The comparison between the unawareness and unawareness groups was summarized in Table 2.

Table 2. EEG and MRI findings, and epilepsy classification
 Unawareness group (n = 31)Awareness group (n = 103)
  1. ap < 0.001; bp = 0.049 (bilateral/unilateral/normal); cp = 0.093, (temporal lobe epilepsy/extratemporal lobe epilepsy and nonlocalized epilepsy).

  2. IEDs, interictal epileptiform discharges; NL, nonlateralized; temporal plus, IEDs or structural lesions involving temporal and extratemporal regions additionally; MTR, mesial temporal region.

EEG findings
Lateralization of IEDs (%)a
 Bilateral15 (48.4)13 (12.6)
 Left 9 (29.0)24 (23.3)
 Right 5 (16.1)20 (19.4)
Location of IEDs (%)
 Temporal24 (77.4)43 (41.7)
 Temporal plus 5 (16.1)7 (6.8)
 Extratemporal0 (0.0)7 (6.8)
Absence of IEDs2 (6.5)46 (44.7)
MRI findings
Location (%)
 Temporal (n = 56)11 (35.5)45 (43.7)
 Temporal plus (n = 13)3 (9.7)10 (9.7) 
 Extratemporal (n = 13)3 (9.7)10 (9.7) 
 Lesion with MTR involvement (n = 53) 9 (29.0)44 (42.7)
 Lesion without MTR involvement (n = 29) 8 (25.8)21 (20.4)
Lateralization (%)b
 Bilateral (n = 10) 5 (16.1)5 (4.9)
 Unilateral 7 (22.6)27 (26.2)
 Left (n = 34)
 Right (n = 38) 5 (16.1)33 (32.0)
Normal (n = 52)14 (45.2)38 (36.9)
Epilepsy classification
Location (%)c
 Temporal lobe26 (83.9)72 (69.9)
 Extratemporal3 (9.7)17 (16.5)
 Nonlocalized2 (6.5)14 (13.6)
Lateralization (%)
 Left14 (45.2)43 (41.7)
 Right 9 (29.0)42 (40.8)
 Nonlateralized 8 (25.8)18 (17.5)

IEDs (93.5% vs. 55.3%) and bilateral IEDs (48.4% vs. 12.6%) were more frequently detected in the unawareness group than in the awareness group (p < 0.001). IEDs localized mainly to the temporal region in both groups (82.8% in the unawareness group and 75.4% in the awareness group).

MRI detected various lesions in 17 (54.8%) patients of the unawareness group and 65 (63.1%) patients of the awareness group (Table 3). Ten (12.2%) of 82 patients with MRI lesions had multiple lesions involving both hemispheres. The bilateral presence of lesions was more frequent in the unawareness group than in the awareness group (16.1% vs. 4.9%, p = 0.049). Structural lesions involving the mesial temporal region and normal finding were found in 29.0% and 45.2% of the unawareness group, respectively, compared with 42.7% and 36.9% of the awareness group. This difference was not statistically significant (p = 0.392).

Table 3. MRI findings
 Unawareness group (n = 31)Awareness group (n = 103)Total
  1. Dual pathology (hippocampal sclerosis plus): cerebromalacia (n = 3), calcification, cortical dysplasia, and granuloma; Atrophic change: cerebromalacia (n = 13), atrophy (n = 2), calcification (n = 2), cyst (n = 2), infarction (n = 2), and cerebromalacia with calcification; Other: tuberous sclerosis (n = 2), cortical dysplasia, and schizencephaly.

Hippocampal sclerosis (%) 7 (22.6)30 (29.1)37 (27.6)
Dual pathology (%)0 (0.0)6 (5.8)6 (4.5)
Atrophic change (%) 6 (19.4)16 (15.5)22 (16.4)
Cavernous hemangioma (%)2 (6.5)5 (4.9)7 (5.2)
Tumor (%)1 (3.2)5 (4.9)6 (4.5)
Other (%)1 (3.2)3 (2.9)4 (3.0)
Normal (%)14 (45.2)38 (36.9)52 (38.8)

Ninety-eight (73.1%) patients were classified as having temporal lobe epilepsy (TLE) based on the lobar localization of epileptogenic foci as determined by the correlation of ictal semiology, EEG, and MRI features. TLE had a slightly higher frequency in the unawareness group (83.9%) than in the awareness group (69.9%), an observation that is not statistically different (p = 0.093). There was no difference between the unawareness and awareness groups in the distribution of lateralization in unilateral IEDs, unilateral lesions on MRI, or epilepsy.

Of four significant variables (age at visit, age at onset, and lateralization of IEDs and MRI lesions) on univariate analyses, only one variable emerged to be significant in the used logistic regression; patients with bilateral IEDs were more likely to be unaware of CPSs than those with absence of IEDs (odds ratio = 33.67; 95% CI = 5.99, 189.39; p < 0.001).

In the subgroup analysis of the patients classified to having TLE the presence of bilateral IEDs and MRI lesions were significantly more frequent in the unawareness group than in the awareness group (data not shown). Structural lesions involving the mesial temporal region on MRI were less frequent (34.6% vs. 61.1%) and normal MRI was more frequent (46.2% vs. 23.6%) in the unawareness group, which has reached to statistical significance (p = 0.046).

Follow-up of patients with complete unawareness of CPSs

During the follow-up (mean of 26 months, ranging from 4 to 55 months), 6 (26.1%) of 23 patients with previously complete unawareness had become aware of CPSs and two of these patients had even experienced the emergence of de novo auras (anxiety and epigastric uprising sensation) although they all had shown incomplete awareness. Of these six patients, four had not been treated with AEDs at their initial visit. Eight patients did not have any further seizures during the follow-up, and the remaining nine patients had experienced CPSs and remained unaware of their seizure occurrence.


Prior studies using various methods of continuous EEG monitoring have shown that a significant proportion of patients were not aware of their seizures (Gotman, 1990; Blum et al., 1996; Tatum et al., 2001). Gotman (1990) reported that 69 (39%) of 179 scalp-recorded seizures detected by an automatic seizure recognition program were not detected by the patient. Tatum et al. (2001) found that 11 (23.4%) of 47 computer-assisted ambulatory EEG recordings contained only partial seizures that were unidentified by the patient, and 7 (14.9%) recordings detected both identified and unidentified partial seizures. Blum et al. (1996) used a video-EEG monitoring unit to directly test self-perception of seizures and found that only 6 (26%) of 23 patients were always aware of their seizures including CPSs and secondarily generalized seizures, 7 (30%) were never aware of any seizures, and 10 (43%) had incomplete awareness of their seizures. The degree of seizure awareness was lowest for patients with temporal lobe foci, especially on the left side. This underreporting of seizures by patients could complicate the results for efficacy in AED trials. Although seizure-free rates could be overestimated in the AED trials, the increased vigilance of patients and their family members after randomization to a new treatment could potentially increase seizure identification and decrease the actual difference from the baseline rate (Gilliam, 2003).

Major limitations of our study are a lack of documentation of unawareness by video-EEG or ambulatory EEG monitoring, and uncertainty of lobar localization and hemispheric lateralization of epileptogenic zone, which were obviously related to the outpatient evaluation. In fact, these limitations might explain a lower frequency of our patients showing unawareness of their CPSs compared with those reported in previous studies (Gotman, 1990; Blum et al., 1996; Tatum et al., 2001), and no significant difference between the unawareness and awareness groups in the distribution of location and lateralization in epilepsy. Therefore, we excluded secondarily generalized tonic–clonic seizures (SGTCSs) from consideration, because SGTCSs are obviously associated with various environmental cues and postictal symptoms that might affect the evaluation for patient's awareness of SGTCSs on an outpatient basis. The purpose of this study was to investigate the incidence of unawareness of CPSs and related clinical variables in a relatively unselected outpatient population, in which the only method of detecting self-perception of seizures is a thorough interview with the patient and family members. The incidence of unawareness may be affected considerably by the family members' attention to the patients, and some patients with awareness may be found to have incomplete unawareness of CPSs if they are monitored by EEG. Thus, the number of patients with incomplete unawareness in this study might be underestimated.

Lux et al. (2002) examined the constituent functions of ictal consciousness including orientation behavior, expressive and receptive speech, and postictal memory, and reported that consciousness (impairment of all examined functions) was more frequently impaired in patients with bitemporal seizure activity compared with those with unilateral temporal or frontal seizure activity. We found that bilateral IEDs and lesions significantly contributed to the unawareness of CPS. Bitemporal IEDs are known to be more often associated with nonlateralized or bilateral ictal activity, bilateral independent seizure onset, asynchrony of ictal activity, and switch of ictal activity, suggesting that bitemporal excitability promotes the rapid propagation of ictal discharges to the bitemporal regions (Steinhoff et al., 1995). Lack of aura experience that should be a prerequisite for unawareness of CPSs, strongly correlates with indicators of bitemporal dysfunction, such as bitemporal IEDs and bitemporal ictal propagation in scalp EEG, and absence of lateralized mesial temporal sclerosis on MRI in patients with mesial TLE (Schulz et al., 2001). Neocortical TLE presents with more frequent and rapid bilateral ictal EEG changes or impairment of consciousness compared to mesial TLE (O'Brien et al., 1996; Maillard et al., 2004; Chabardes et al., 2005), which could be expected in our TLE patients with unawareness showing more frequent normal finding and less frequent structural lesions involving the mesial temporal region on MRI. Bitemporal IEDs as well as bilateral lesions might inhibit interictal memory function not only ipsilateral but also contralateral to the seizure onset. The Wada test often fails to lateralize the epileptogenic region in patients with bitemporal IEDs (Benbadis et al., 1995). Asymmetry of memory fMRI activation in the mesial temporal structures was found less frequently in mesial TLE patients with bitemporal IEDs than in those with unitemporal IEDs (Janszky et al., 2004).

Therefore, bilateral temporal dysfunction and rapid spread of ictal discharges to the contralateral hemisphere related to bilateral independent IEDs and bilateral presence of lesions, may result in prompt and complete loss of ictal consciousness and memory dysfunction that prevent the awareness of CPSs. It was interesting to find that some patients with complete unawareness of CPSs had gained the seizure awareness after institution of AED therapy. It is likely that AED therapy shortened seizure duration, slowed propagation of ictal discharge, and limited the extent of ictal discharge spreading, thereby limiting the unawareness of seizures. This phenomenon might also explain the higher frequency of untreated status at the initial visit in patients with complete unawareness in this study as well as much higher incidence of seizure unawareness in a study by Blum et al. (1996) who directly investigated seizure awareness by using video-EEG monitoring, probably during rapid AED tapering or after AED withdrawal.

This study suggests that a significant portion of relatively unselected patients with CPSs are unaware of their CPSs. Particular attention should be give to older patients with bilateral IEDs or lesions. Investigators should cooperate thoroughly with these patients' family members and friends to monitor seizure occurrence.