1

Timing of seizure occurrence and response of

medical staff in video-EEG monitoring for

drug-resistant epilepsy

Alief Dhuha, Takehiko Konno, Hidenori Yokota, Keiji Oguro, Eiju Watanabe, Kensuke Kawai

Purpose: The purpose of this study was to determine the optimal duration of long-term video-

EEG monitoring (VEEG) in patients with epilepsy. The response time of medical staff to seizures

was also evaluated from the viewpoint of safety of the monitoring.

Methods: We estimated the optimal duration of VEEG from the seizure onset pattern. We retro-

spectively investigated all VEEG sessions performed in our department during the period between

June 2005 and July 2016. Sessions with no seizures and with only non-epileptic seizures were ex-

cluded. Using 91 sessions from 69 patients, information on the onset time and response time of

medical staff to seizures was collected.

Results: The median duration from the start of VEEG to the first seizure was 2 days. Seventy-seven

percent of first seizures occurred within 3 days of VEEG. The median duration from the first sei-

zure to the third seizure was 2 days. Eighty percent of third seizures occurred within 3 days of the

first seizure. There was no significant diurnal distribution of seizure occurrence. Medical staff did

not respond to 20% of generalized seizures and 69% of focal seizures. The overlooking of general-

ized seizures occurred mainly during the hours of 1-2 pm and 8-9 pm but there was no significant

diurnal pattern in overlooking generalized and focal seizures.

Conclusion：Based on these findings, we suggest that VEEG can be terminated when no seizure

occurs within 4 days after onset. In our VEEG protocol, in which all antiepileptic drugs were dis-

continued before the start of a session, there was no diurnal pattern of seizure occurrence. This is

the first study investigating the diurnal pattern of overlooking seizures by medical staff during

VEEG. Since there was no diurnal pattern to the overlooking, medical staff should pay equal, 24-

hour attention to patients on VEEG.

Corresponding author: Dr. Kensuke Kawai

Department of Neurosurgery, Jichi Medical University, 3311-1 Yakushiji, Simotsuke, Tochigi 329-0498, Japan.

Tel: +81-285-58-7373, Fax: +81-285-44-5147, E-mail: kenkawai-tky@jichi.ac.jp

Epilepsy & Seizure Journal of Japan Epilepsy Society

Vol. 9 No.1 (2017) pp. 1-10

Department of Neurosurgery, Jichi Medical University, Simotsuke, Tochigi, Japan

Abstract

Received: January 7, 2017, Accepted: March 4, 2017

Key words: Epilepsy, seizure, video-EEG, long-term monitoring, presurgical evaluation

Original Article

2

Introduction

Video-EEG (VEEG) monitoring is recom-

mended as the gold standard for diagnosis of

epilepsy, classification of seizures/epilepsy,

and pre-surgical evaluation [1–7]. The opti-

mal duration of VEEG monitoring has not

been determined but it was found to be be-

tween 3 to 6 days in most reported studies to

capture more than one seizure [1, 8–10]. The

optimal protocol for VEEG monitoring, in-

cluding the duration of monitoring and time

of discontinuation of anti-epileptic drugs

(AEDs), also has not been determined. Fur-

thermore, occurrence of seizures, particularly

generalized seizures, during VEEG monitor-

ing may cause major adverse events includ-

ing physical injuries and even death from sei-

zures [3, 7, 9, 11–15]. Medical staff plays

important roles in preventing these events

during VEEG monitoring [11]. Our study

aimed to determine the optimal duration of

VEEG monitoring in patients with epilepsy

by retrospectively analyzing the timing and

pattern of seizure occurrence during VEEG

monitoring. From the view point of safety of

the monitoring, the responses of medical staff

to seizures were also evaluated.

Methods Patients and VEEG data

The data for this study were obtained from

a database of digital EEG and movie files

during VEEG monitoring in the Department

of Neurosurgery at Jichi Medical University.

A list of patients who underwent VEEG dur-

ing the period starting from June 2005 was

constructed by the authors (E.W., K.O.,

H.Y.), which included patients’ names, ID

numbers, and presence or absence of im-

portant events (prominent epileptiform dis-

charges, seizures and suspected seizures) dur-

ing VEEG sessions. The corresponding digi-

tal data of VEEG for 2 h which included each

event were available for review. Not all digi-

tal data throughout each session were availa-

ble because of the limited capacity of hard

disks. VEEG monitoring was performed in

patients with drug-resistant epilepsy to deter-

mine indication for surgical treatment and/or

to localize epileptic focus in presurgical

workup. Three patients who had generalized

epilepsy were excluded from this study.

The list included 120 sessions obtained

from 115 patients during a period of more

than 10 years between June 2005 and July

2016. All the corresponding digital data for

each event were reviewed by one of the au-

thors (A.D.) in terms of whether the event

was a true epileptic seizure based upon EEG

and clinical symptoms. In the case of true ep-

ileptic seizures, they were classified into fo-

cal or generalized seizures. The starting and

ending times for each seizure were deter-

mined. Sessions with only psychogenic non-

epileptic seizures (N = 8) and sessions with-

out any seizures (N = 21) were excluded.

Twenty-five seizures were excluded because

the corresponding 2-h digital EEG file or

movie file was missing. Consequently, 91

sessions with 650 true epileptic seizures ob-

tained from 69 patients were analyzed in the

present study. Four patients had two scalp

EEG sessions and one patient had three scalp

EEG sessions. Eighteen patients had one in-

tracranial EEG session and two patients had

two intracranial EEG sessions. All patients

who underwent intracranial EEG sessions had

preceding scalp EEG sessions.

Video-EEG monitoring for drug-resistant epilepsy Alief Dhuha, et al.

3

VEEG monitoring

A single monitoring room equipped with a

128-channel EEG system and a video camera

(Nihon Kohden® 1100, Tokyo, Japan) was

used. A mobile toilet was placed in the room.

Patients had their meals in the room. On the

first day, scalp EEG was prepared using the

international 10-20 system. Sphenoidal elec-

trodes were used when needed (N = 38). For

intracranial studies, subdural electrodes and/

or depth electrodes were used as indicated.

Patients were instructed to push a nurse-call

button or an EEG-marked button when they

had seizures. There was no accompanying

person during monitoring except visitors. A

video feed was streamed for 24 h in the nurs-

es’ station, but it was not continuously moni-

tored by an assigned observer.

In principle, all AEDs were discontinued on

the day before the start of VEEG monitoring,

all of which were resumed immediately after

termination of VEEG. However, in 14 ses-

sions, some AEDs were continued because

the patients had daily seizures even under

AEDs, and these patients were excluded from

the study of the timing of seizures. Recorded

VEEG data were checked twice a day by the

authors (K.O., H.Y. and A.D.). Important

events were marked in the digital data and

were listed up. There were no strict criteria

for terminating the session, but in principle

VEEG was continued until more than 5 sei-

zures were captured or otherwise for 1 week.

There were no strict criteria for resuming

AEDs either, but in principle pre-monitoring

AEDs were resumed when three seizures

were captured for analysis.

Data acquisition and analysis

For the study on the timing of seizures, we

used 571 seizures in 77 sessions from 58 pa-

tients, after excluding 14 sessions in which

AEDs were not discontinued before the start

of VEEG sessions. The starting time of every

seizure was obtained and expressed as the 24-

hour clock time of a given day. The time of

each day was defined as from 00:00 to 24:00,

and not as the time elapsed from the start of

the session. For a given session, the number

of seizures per hour of each day was counted.

The number of days from the start of a ses-

sion to the first seizure and the number of

days from the first seizure to the third seizure

were determined for each session. Then we

counted the number of sessions during which

the first seizure occurred on each day. Simi-

larly, we counted the number of sessions dur-

ing which the third seizure occurred on each

day from the day of the first seizure. Because

11 patients experienced only one or two sei-

zures during monitoring sessions, the latter

analysis was done using 66 sessions from pa-

tients who had three or more seizures.

To test the hypothesis that patients with a

higher seizure frequency had the first seizure

earlier, we analyzed the association between

seizure frequency and the day of the first sei-

zure. The seizure frequency of each session

was calculated by dividing the total number

of seizures during the session by the number

of days of session. Similarly, we tested the

hypothesis that patients with a higher seizure

frequency had the third seizure after a shorter

duration from the first seizure.

To study the diurnal distribution of seizure

occurrence, we used the above-mentioned

data on the total number of seizures per hour

Epilepsy & Seizure Vol. 9 No. 1 (2017)

4

of each day in all 91 sessions. We used two

methods to evaluate the diurnal distribution.

First, we adopted a method that was used in a

previous study [16]. The 24-h day was divid-

ed into four bins. The summated numbers of

seizures in all 91 sessions in the four bins

were compared. Since the effects of inter-

patient differences in seizure frequency and

session length cannot be avoided in this

method, we compared the diurnal rates of sei-

zure occurrence in each patient. For each pa-

tient, we summated the number of seizures in

all sessions for each hour of the day, and de-

termined the rate of seizure occurrence in the

given bins of that day.

Whether any nurse and/or doctor appeared

in the movie during seizure was also checked.

When they were observed on the video, the

seizure was regarded as response positive.

Response negative was defined as when nei-

ther a nurse nor a doctor appeared in the

movie from the start to finish of a seizure.

The diurnal distribution of response to sei-

zures was evaluated. For each time bin, the

summated numbers of response-positive sei-

zures and response-negative seizures were

determined using all session data. The null

hypothesis was that the numbers are equally

distributed in all time bins.

Statistical analysis

Student’s t-test was used for the compari-

son of seizure frequency between the patient

groups in which patients had the first seizure

before and after a given day, and between the

patient groups in which patients had the third

seizure before and after a given number of

days following the first seizure. Repeated

ANOVA was used for the comparison of sei-

zure occurrence during four time bins in 24 h.

The Wilcoxon sign rank test was used for

comparison of seizure occurrence during two

bins in 24 h. IBM® SPSS® 20 was used for

the above-mentioned statistical tests.

Results Patients, VEEG sessions and seizures

We retrospectively analyzed all VEEG ses-

sions in our department for an approximately

10-year period. There were 91 sessions ob-

tained from 69 patients. The median age of

the patients was 31 years (range, 8 to 63

years). Thirty-nine of 69 were males. The

median duration of VEEG sessions was 5

days (range, 1 to 13 days). Six hundred and

fifty seizures were captured in 91 sessions of

the 69 patients. One hundred and seventy-

three were generalized seizures and 477 were

focal seizures. The median number of all sei-

zures in a session was 6 (range, 1 to 38). The

median numbers of generalized and focal sei-

zures were 1 (range, 0 to 22) and 4 (range, 0

to 23), respectively.

Timing of seizure occurrence

Figure 1A demonstrates the day of occur-

rence of the first seizure. The peak number of

first seizures was observed on the second day

and the number gradually decreased thereaf-

ter. The median days from the start of VEEG

to the first seizure was 2 (range, 1 to 10).

Seventy-seven percent of seizures had oc-

curred by the third day. Figure 1B demon-

strates the day of occurrence of the third sei-

zure after the first seizure. The peak number

of the third seizure was again observed on the

second day and the number decreased abrupt-

ly thereafter. The median number of days

from the first seizure to the third seizure was

Alief Dhuha, et al. Video-EEG monitoring for drug-resistant epilepsy

5

2 (range, 1 to 8). Eighty percent of seizures

had occurred by the third day after the first

seizure.

To provide an overview of the association

between the timing of the first seizure and

seizure frequency in a given session, all ses-

sions are plotted in Fig. 2A. The first seizure

tended to occur earlier in sessions with higher

seizure frequencies. To confirm this tenden-

cy, we compared the seizure frequency be-

tween sessions with the first seizure before

and after a given day. We found that seizure

frequency was significantly higher in sessions

with the first seizure occurring before the sec-

ond or third day than in sessions with the first

seizure occurring after those days.

A similar analysis was performed for the

day of occurrence of the third seizure (Fig.

2B). We found that seizure frequency was

significantly higher in sessions with the third

seizure occurring before the second or third

day after the first seizure than in sessions

with the third seizure occurring after those

days.

Epilepsy & Seizure Vol. 9 No. 1 (2017)

Figure 1. Times of occurrence of the fir st seizure and the third seizure. A. Day of occurrence of the fir st seizure. Line graph shows the cumulative percentage of first seizures on each day of monitoring. Bar graph shows the number of first seizures that occurred on each day of monitoring. B. Day of occurrence of the third seizures. Line graph shows the cumulative percentage of third seizures on each day after the first seizure. Bar graph shows the number of third seizures that occurred on each day after the first seizure.

Figure 2. Association between seizure frequency and timing of the fir st seizure (A) and third seizure (B). All sessions were plotted using the days of the first and third seizures (X axis) and mean seizure frequencies per day (Y axis).

A B

A B

6

Diurnal pattern of seizure occurrence

While there were two peaks; between 6 am

and 7 am and between 6 pm and 7 pm, in the

occurrence of focal seizures, there was no

apparent peak in the occurrence of general-

ized seizures (Fig. 3). Reflecting the peaks in

focal seizures, the total number of seizures

demonstrated two peaks in the same hours.

Although seizure occurrence was apparently

higher from midnight to early morning than

during daytime, we did not detect a statisti-

cally significant diurnal pattern of the occur-

rence of total seizures, generalized seizures or

focal seizures.

Response of medical staff

Twenty percent of 173 generalized seizures

and 69% of 477 focal seizures were over-

looked by medical staff. Figure 4A demon-

strates the diurnal distribution of total and

overlooked generalized seizures. Figure 4B

demonstrates the diurnal distribution of the

rate of overlooking generalized seizures.

While there were apparently peaks between 1

pm and 2 pm and between 8 pm and 9 pm,

the rate varied prominently during a 24-h pe-

riod. We did not detect a significant diurnal

pattern in overlooking of generalized sei-

Alief Dhuha, et al. Video-EEG monitoring for drug-resistant epilepsy

Figure 3. Diurnal pattern of seizure occurrence. The summated numbers of seizures of all sessions during each hour are demonstrated. Triangular, cir-cle, and square marks are total number of seizures, number of generalized seizures, and number of focal seizures, respectively. ‘x’ denotes hour x to x+1; for example, hour 0 denotes 0:00 to 1:00.

Figure 4. Diurnal pattern of medical staff re-sponse to generalized seizure. A. The numbers of generalized seizure and overlooked generalized sei-zure during each hour. B. The rate (%) of overlook-ing of generalized seizure during each hour.

A

B

A

B

Figure 5. Diurnal pattern of medical staff re-sponse to focal seizure. A. The numbers of focal seizure and overlooked focal seizure during each hour. B. The rate (%) of overlooking of focal seizure during each hour.

7

zures. The lack of diurnal pattern in over-

looking was clearer for focal seizures (Fig.

5A, B). Again, we did not detect a significant

diurnal pattern in overlooking of focal sei-

zures. Another significant finding was that

the overlooking rates were much higher for

focal seizures than generalized seizures over

24 h.

Discussion The optimal duration of VEEG monitoring

has not been determined but most studies re-

ported between 3 and 6 days to capture more

than one seizure [1, 9, 10]. Capturing one sei-

zure may be enough to rule out non-epileptic

seizures or to make a diagnosis of epilepsy,

but is not regarded sufficient for presurgical

evaluation. Many researchers have recom-

mended that at least three seizures are re-

quired to properly locate the epileptic region

[8, 9, 14, 17, 18]. Therefore, we investigated

the times of occurrence of the first seizure

and third seizure. The occurrence of the first

seizure declined steeply after 5 days. Seventy

-seven percent of first seizures occurred by

the third day. Eighty percent of third seizures

occurred by the third day after the first sei-

zure. Eighty percent of third seizures oc-

curred by the fifth day of VEEG monitoring.

Knowledge of these patterns of occurrence is

useful to avoid meaningless continuation of

VEEG and to save medical resources.

VEEG has been regarded as the gold stand-

ard for diagnosis and classification of epilep-

sy and presurgical evaluation. Therefore,

nearly all previous retrospective studies on

the usefulness or optimal duration of VEEG

monitoring included a variety of patients with

various purposes for performing VEEG [8,

10]. However, inclusion of diverse patients

with various purposes of VEEG may increase

the variability of the time of seizure occur-

rence. Alving and Beniczky [8] reported that

the duration of successful VEEG monitoring

was significantly longer in the presurgical

group (mean: 3.5 days) than in the diagnostic

and classification groups (2.4 and 2.3 days,

respectively). More recently, Hupalo et al.

[10] reported that VEEG should last at least

72 h in patients with refractory epilepsy,

while most of the cases with psychogenic non

-epileptic seizures could be diagnosed after

48 h. A non-uniform regimen of AED dose

reduction or discontinuation during VEEG

monitoring in previous studies also may have

resulted in greater variability of the time of

seizure occurrence [10, 19]. Our study was

unique and strong in that the patient cohort

was considerably uniform because VEEG

was performed solely for presurgical evalua-

tion in patients with drug-resistant partial epi-

lepsy, and AED was discontinued in all pa-

tients one day before the start of VEEG.

To taper or withdraw AEDs before VEEG

monitoring facilitates recording of seizures

sooner or in larger numbers [10, 20]. Our

findings on the association between the time

of occurrence and the frequency of seizures

support this finding. In one study in which

60% of patients had AEDs reduced by half

but none discontinued AEDs totally, only

15% of epileptic seizures occurred within the

first 24 h of long-term VEEG monitoring

(53% and 32% on the second and third day,

respectively) [10]. In contrast, in our study,

seizure occurrence was higher from the be-

ginning of monitoring, considering that we

started VEEG in the afternoon of the first day

Epilepsy & Seizure Vol. 9 No. 1 (2017)

8

and that the first day represented less than 24

h.

Importantly, tapering or withdrawal of

AEDs has been believed to be potentially

dangerous, as it carries a 3% risk of status

epilepticus or cluster seizures [4, 13, 21, 22].

Therefore, most epilepsy centers gradually

taper or partially discontinue AEDs [10, 19].

There were no previous studies in which all

AEDs were discontinued before the VEEG

admission. Interestingly, we and a number of

other authors observed no complications re-

lated to discontinuation or dose reduction of

AEDs [2, 9, 10, 18, 19, 21–25]. However, it

is too early to conclude that we can discontin-

ue all AEDs from our single series of VEEG

monitoring. It is highly probable that frequen-

cy of seizure and risk of status epilepticus are

dependent on pre-VEEG seizure frequency,

semiology of each patient, and medication

regimen. There is also the concern that dis-

continuing all AEDs may increase the epilep-

tic activity of various brain regions too exten-

sively, making it difficult to properly localize

the true epileptic focus to be surgically treat-

ed.

While seizure prevalence associated with

the wake-sleep cycle has been studied exten-

sively, studies on the diurnal pattern of sei-

zure occurrence are limited. Hofstra et al.

[16, 26] reported that more seizures were ob-

served from 11:00 to 17:00, and fewer sei-

zures from 23:00 to 05:00. They suggested

that certain types of seizures have a strong

tendency to occur in a truly diurnal pattern.

However, they tested the diurnal pattern only

by comparing the total number of seizures

from all patients for a 5-year period in four

bins of 24 h. We did not detect any signifi-

cant diurnal pattern by using more extensive

testing applied to a more homogeneous pa-

tient group. Therefore, at least in partial epi-

lepsy, it is unlikely that seizure occurrence

has a clear diurnal pattern.

Our study is the first to examine the diurnal

pattern of overlooking seizures by medical

staff. In our center, there are three meal times

in a day; 7:30 for breakfast, 11:30 for lunch

and 18:30 for dinner. Nurses are busiest at

those times. Interestingly, the peaks of over-

looking seizures did not coincide with those

busiest hours. It is noteworthy that 20% of

generalized seizures and 69% of focal sei-

zures were overlooked by medical staff.

Spanaki et al. [24] reported that focal seizures

had a tendency to be missed by medical staff

during VEEG sessions due to difficulties in

visual detection. Improvement in detection

measures for generalized and focal seizures

may improve the safety of VEEG and save

medical resources by avoiding futilely

lengthy VEEG.

References [1] Chang BS, Ives JR, Schomer DL, Dris-

lane FW. Outpatient EEG monitoring in

the presurgical evaluation of patients

with refractory temporal lobe epilepsy. J

Clin Neurophysiol 2002; 19(2): 152–6.

[2] Bauer J, Ben-Menachem E, Kramer G,

Fryze W, Da Silva S, Kasteleijn-Nolst

Trenite DG. Levetiracetam: a long-term

follow-up study of efficacy and safety.

Acta Neurol Scand 2006; 114(3): 169–

76.

[3] Friedman DE, Hirsch LJ. How long does

it take to make an accurate diagnosis in

an epilepsy monitoring unit? J Clin Neu-

rophysiol 2009; 26(4): 213–7.

Alief Dhuha, et al. Video-EEG monitoring for drug-resistant epilepsy

9

[4] Dobesberger J, Walser G, Unterberger I,

Seppi K, Kuchukhidze G, Larch J, et al.

Video-EEG monitoring: Safety and ad-

verse events in 507 consecutive patients.

Epilepsia 2011; 52(3): 443–52.

[5] Di Gennaro G, Picardi A, Sparano A,

Mascia A, Meldolesi GN, Grammaldo

LG, et al. Seizure clusters and adverse

events during pre-surgical video-EEG

monitoring with a slow anti-epileptic

drug (AED) taper. Clin Neurophysiol

2012; 123(3): 486–8.

[6] Beghi E, Giussani G, Grosso S, Iudice

A, Neve A La, Pisani F, et al. Withdraw-

al of antiepileptic drugs: Guidelines of

the Italian League Against Epilepsy. Epi-

lepsia 2013; 54(SUPPL.7): 2–12.

[7] Fisher RS, Acevedo C, Arzimanoglou A,

Bogacz A, Cross JH, Elger CE, et al.

ILAE Official Report: A practical clini-

cal definition of epilepsy. Epilepsia

2014; 55(4): 475–82.

[8] Alving J, Beniczky S. Diagnostic useful-

ness and duration of the inpatient long-

term video-EEG monitoring: Findings in

patients extensively investigated before

the monitoring. Seizure 2009; 18(7): 470

–3.

[9] Gazzola DM, Thawani S, Agbe-Davies

O, Carlson C. Epilepsy monitoring unit

length of stay. Epilepsy Behav 2016; 58:

102–5. Available from: http://

linkinghub.elsevier.com/retrieve/pii/

S1525505016000986

[10] Hupalo M, Smigielski JW, Jaskolski DJ.

Optimal time of duration of a long-term

video-EEG monitoring in paroxysmal

events - A retrospective analysis of 282

sessions in 202 patients. Neurol Neuro-

chir Pol 2016; 50(5): 331–5. Available

from: http://www.ncbi.nlm.nih.gov/

pubmed/27591057

[11] Foley J, Oates J, Mack C, Fox C. Im-

proving the epilepsy service: The role of

the specialist nurse. Seizure 2000; 9(1):

36–42.

[12] Ghougassian DF, D’Souza W, Cook

MJ, O’Brien TJ. Evaluating the utility

of inpatient video-EEG monitoring. Epi-

lepsia 2004; 45(8): 928–32.

[13] Atkinson M, Hari K, Schaefer K, Shah

A. Improving safety outcomes in the

epilepsy monitoring unit. Seizure 2012;

21(2): 124–7. Available from: http://

dx.doi.org/10.1016/

j.seizure.2011.10.004

[14] Henning O, Baftiu A, Johannessen SI,

Landmark CJ. Withdrawal of antiepilep-

tic drugs during presurgical video-EEG

monitoring: An observational study for

evaluation of current practice at a refer-

ral center for epilepsy. Acta Neurol

Scand 2014; 129(4): 243–51.

[15] Michel V, Mazzola L, Lemesle M, Ver-

cueil L. Long-term EEG in adults: Sleep

-deprived EEG (SDE), ambulatory EEG

(Amb-EEG) and long-term video-EEG

recording (LTVER). Neurophysiologie

Clinique 2015; 45: 47–64.

[16] Hofstra WA, Spetgens WPJ, Leijten

FSS, van Rijen PC, Gosselaar P, van der

Palen J, et al. Diurnal rhythms in sei-

zures detected by intracranial electro-

corticographic monitoring: An observa-

tional study. Epilepsy Behav 2009; 14

(4): 617–21. Available from: http://

dx.doi.org/10.1016/j.yebeh.2009.01.020

[17] Haut SR, Swick C, Freeman K, Spencer

S. Seizure clustering during epilepsy

monitoring. Epilepsia 2002; 43(7): 711–

Epilepsy & Seizure Vol. 9 No. 1 (2017)

10

5.

[18] Ferastraoaru V, Schulze-Bonhage A,

Lipton RB, Dümpelmann M, Legatt

AD, Blumberg J, et al. Termination of

seizure clusters is related to the duration

of focal seizures. Epilepsia 2016; 57(6):

889–95.

[19] Hui ACF, Kwan P, Leung TW, Soo Y,

Mok VCT, Wong LKS. Diagnostic val-

ue and safety of long-term video-EEG

monitoring. Hong Kong Med J 2007;

13: 228–30.

[20] Velis D, Plouin P, Gotman J, Da Silva

FL. Recommendations regarding the

requirements and applications for long-

term recordings in epilepsy. Epilepsia

2007; 48(2): 379–84.

[21] Rose AB, McCabe PH, Gilliam FG,

Smith BJ, Boggs JG, Ficker DM, et al.

Occurrence of seizure clusters and status

epilepticus during inpatient video-EEG

monitoring. Neurology 2003; 60(6): 975

–8.

[22] Dobesberger J, Ristić AJ, Walser G,

Kuchukhidze G, Unterberger I, Höfler J,

et al. Duration of focal complex, sec-

ondarily generalized tonic-clonic, and

primarily generalized tonic-clonic sei-

zures - A video-EEG analysis. Epilepsy

Behav 2015; 49: 111–7.

[23] Noe KH, Drazkowski JF. Safety of long-

term video-electroencephalographic

monitoring for evaluation of epilepsy.

Mayo Clin Proc 2009; 84(6): 495–500.

[24] Spanaki M V, Mccloskey C, Remedio V,

Budzyn D, Guanio J, Monroe T, et al.

Developing a culture of safety in the epi-

lepsy monitoring unit: A retrospective

study of safety outcomes. Epilepsy Be-

hav 2012; 25(2): 185–8. Available

from: http://dx.doi.org/10.1016/

j.yebeh.2012.06.028

[25] Fahoum F, Omer N, Kipervasser S, Bar-

Adon T, Neufeld M. Safety in the epi-

lepsy monitoring unit: A retrospective

study of 524 consecutive admissions.

Epilepsy Behav 2016; 61: 162–7. Avail-

able from: http://

linkinghub.elsevier.com/retrieve/pii/

S1525505016301445

[26] Hofstra WA, Grootemarsink BE, Dieker

R, Van Der Palen J, De Weerd AW.

Temporal distribution of clinical sei-

zures over the 24-h day: A retrospective

observational study in a tertiary epilepsy

clinic. Epilepsia 2009; 50(9): 2019–26.

Alief Dhuha, et al. Video-EEG monitoring for drug-resistant epilepsy