J Neurol
DOI 10.1007/s00415-015-7992-0
ORIGINAL COMMUNICATION
Association of seizure duration and outcome in refractory
status epilepticus
Dominik Madžar1 • Anna Geyer1 • Ruben U. Knappe1 • Stephanie Gollwitzer1 •
Joji B. Kuramatsu1 • Stefan T. Gerner1 • Hajo M. Hamer1 • Hagen B. Huttner1
Received: 18 October 2015 / Revised: 1 December 2015 / Accepted: 8 December 2015
Ó Springer-Verlag Berlin Heidelberg 2015
Abstract The aim of the study was to identify factors
influencing long-term outcome and to evaluate the prog-
nostic power of the Status Epilepticus Severity Score
(STESS) in refractory status epilepticus (RSE). We retro-
spectively extracted data on baseline characteristics, RSE
details, and hospital course including complications from
all patients treated for RSE in our institution between
January 2001 and January 2013. Functional outcome was
assessed using the modified Rankin Scale (mRS) and was
defined as good when either RSE did not lead to functional
decline or when the resulting mRS score was 2 or below.
Seventy-one episodes in 65 patients were analyzed. The
median follow-up time was 12 weeks (IQR 6–35), two
patients were lost to follow-up. Poor functional long-term
outcome was observed in 42/69 (60.9 %) episodes. In-
hospital mortality occurred in 13/71 (18.3 %) episodes.
Multivariable analysis revealed that STESS C 3, longer
RSE duration, and sepsis were independently related to
poor functional long-term outcome. Receiver operating
characteristics (ROC) curve analyses confirmed the cut-off
dichotomization into STESS C 3 and STESS \ 3 for
optimal discrimination between good and poor outcome
(AUC = 0.671, p = 0.002, YI = 0.368, NPV = 0.607,
PPV = 0.756) and revealed an RSE duration of 10 days as
a significant cut-off point associated with outcome
Electronic supplementary material The online version of this
article (doi:10.1007/s00415-015-7992-0) contains supplementary
material, which is available to authorized users.
& Dominik Madžar
dominik.madzar@uk-erlangen.de
1
Department of Neurology, University of Erlangen-Nuremberg,
Schwabachanlage 6, 91054 Erlangen, Germany
(AUC = 0.712, p = 0.012, YI = 0.310; NPV = 0.545,
PPV = 0.750). In conclusion, STESS and RSE duration
represent relevant scores and parameters impacting long-
term outcome after RSE. A shorter RSE duration is asso-
ciated with better outcome and, therefore, rapid and ade-
quate treatment for seizure termination should be enforced.
Keywords Refractory status epilepticus
Outcome

Status Epilepticus Severity Score
Seizure duration
Introduction
Despite the lack of a formal definition, status epilepticus
(SE) is generally termed refractory when administration
of at least two properly dosed antiepileptic drugs (AEDs)
fails to terminate seizures [1]. Around 10–40 % of SE
episodes progress into refractory SE (RSE) [2] in which
therapeutic coma induction is frequently required to
achieve seizure control. Compared to SE, RSE has been
associated with worse discharge outcome [3, 4], but only
limited data are available regarding long-term functional
outcome after RSE. While studies consistently report a
poor long-term prognosis for the majority of patients, a
substantial proportion of RSE episodes result in complete
recovery [5–10]. Thus, in light of this wide variety in
outcome and given the lack of approved predictive
parameters, the individual patient’s risk stratification is
difficult [1]. While the Status Epilepticus Severity Score
(STESS) prognosticates survival and return to premorbid
functional status in SE, its validity in RSE is uncertain
and remains to be established [11]. The present study
sought: (a) to identify predictors for long-term outcome
and (b) to evaluate the prognostic power of STESS in
RSE.
123

Methods
Status epilepticus definitions
According to previous investigations, SE was defined as
clinical and/or electroencephalographic evidence of seizure
activity for C5 min or as a series of seizures without
interictal clinical recovery [3, 12–14]. RSE was defined as
SE with persistence of clinical and/or electroencephalo-
graphic seizure activity despite sufficiently dosed treatment
with at least two AEDs [5, 15, 16].
Identification of RSE episodes and inclusion/
exclusion criteria
The medical records of all patients treated for SE in our
institution between January 2001 and January 2013 were
reviewed for episodes meeting the criteria of RSE. When
chart review confirmed the diagnosis of RSE, episodes
were included into the study only if the records contained
sufficient data on RSE characteristics, treatment, and
complications during the hospital stay. For comparability
with previous studies [5, 17], RSE episodes caused by
hypoxic encephalopathy, simple-partial, and absence RSE
episodes were excluded from analysis.
Collection of clinical and RSE-specific data
The following data were extracted from the records: age
at admission, gender, premorbid functional status, history
of seizures, laboratory findings on admission, need for
vasopressors, mechanical ventilation, and duration of
hospital stay. RSE characteristics included duration of
RSE, presumptive etiology of RSE—according to ILAE
criteria [18] including ‘‘potentially fatal’’ etiology as
defined previously [19], number of AEDs administered,
use of anesthetics, and induction of burst suppression.
RSE duration was defined as the time from onset to RSE
termination, with the latter diagnosed either based on
electroencephalography (EEG) results or on documented
clinical improvement clearly indicative of an end of sei-
zure activity [20]. RSE was graded using the Status
Epilepticus Severity Score (STESS) and episodes were
dichotomized into STESS of 3 or higher (STESS C 3) and
STESS of less than three points (STESS \ 3), as proposed
by Rossetti and coworkers [11].
Outcome definition and assessment
Functional outcome was quantified using the modified
Rankin Scale (mRS) [21]. For every RSE episode we
evaluated two outcomes, one at discharge and one long-
123
J Neurol
term outcome. The latter was assessed using the last
available follow-up data documented in our records. Epi-
sodes were defined to have resulted in good outcome when
RSE either did not lead to functional decline (i.e., no
increase in the mRS score relative to baseline) or when the
resulting mRS score after RSE was B2. Otherwise, out-
come was classified as poor. We a priori decided to per-
form subanalyses for (a) patients who died in hospital, as
well as (b) episodes which resulted in complete return to
premorbid neurological status.
Statistical analysis
Statistical analysis was performed using IBM SPSS
Statistics 20.0 (http://www.spss.com). p values of 0.05 or
less were considered statistically significant. All statistical
tests were two-sided. The Pearson-v2 and the Fisher’s-exact
tests were applied to compare frequency distributions of
categorical variables. The Kolmogorov–Smirnov test was
applied to distinguish normal from non-normal distribu-
tion. Normally distributed data are presented as
mean ± standard deviation (SD) (compared using the
Student’s t test). Others are displayed as median and
interquartile range (IQR) (compared using the Mann–
Whitney U test). We calculated forward inclusion multi-
variable logistic regression models to identify parameters
independently associated with outcome (including all sig-
nificant variables upon univariate testing). An exception
was made for the dichotomized STESS which was forced
into models in case of an absent univariate significance.
Variables representing components of STESS were not
entered separately into models. The Hoshmer–Lemeshow
goodness-of-fit test was applied to evaluate regression
models. Where appropriate, receiver operating character-
istics (ROC) curve analyses were performed for cut-off
point determinations.
Results
Demographics and RSE characteristics
We reviewed 445 SE episodes in 330 patients and among
these identified 74 RSE episodes in 68 patients. Three
episodes had to be excluded from the study due to insuf-
ficient documentation of the clinical course. All three were
in patients transferred to our institution for further treat-
ment from another hospital. Therefore, 71 episodes in 65
patients remained for final analysis. The worst seizure
types were complex-partial in 31/71 (43.7 %), generalized
convulsive in 31/71 (43.7 %), and nonconvulsive SE
(NCSE) in coma in 9/71 (12.7 %) episodes. Causes of RSE
were acute symptomatic in 31/71 (43.7 %), cryptogenic in
J Neurol

Table 1 Overview of patient characteristics and parameters associated with long-term outcome
Total cohort Good long-term outcome Poor long-term outcome p value
(n = 71) (27 of 69 episodes) (42 of 69 episodes)

Demographics
Gender, female 48 (67.6) 19 (70.4) 28 (66.7) 0.747
Age on admission (years) 63.8 (±19.2) 54.1 (±22.0) 69.9 (±15.0) 0.002
Age ‡ 65 years 42 (49.2) 11 (40.7) 29 (69.0) 0.020
Premorbid mRS 3 (1–4) 2 (0–4) 3 (1–4) 0.233
Baseline RSE characteristics
NCSE in coma 9 (12.7) 1 (3.7) 8 (19.0) 0.079
Generalized convulsive SE 31 (43.7) 15 (55.6) 14 (33.3) 0.068
Complex-partial SE 31 (43.7) 11 (40.7) 20 (47.6) 0.575
Stuporous or comatose 44 (61.2) 16 (59.3) 26 (61.9) 0.826
Acute symptomatic etiology 31 (43.7) 12 (44.4) 18 (42.9) 0.897
History of seizures 37 (52.1) 19 (70.4) 17 (40.5) 0.015
Potentially fatal etiology 32 (45.1) 9 (33.3) 22 (52.4) 0.121
STESS 3 (2–4) 2 (2–4) 3 (2–5) 0.016
STESS ‡ 3 43 (60.6) 10 (37.0) 31 (73.8) 0.002
Laboratory findings on admission
Leukocyte count (9103/ll) 9.4 (6.9–12.4) 9.4 (8.1–12.9) 9.4 (6.2–12.2) 0.554
C-reactive protein (mg/l) 16.3 (3.2–53.7) 5.2 (1.6–17.7) 25.2 (5.4–56.3) 0.005
Hemoglobin (g/dl) 11.7 (±1.7) 12.0 (±1.8) 11.5 (±1.7) 0.283
Serum sodium (mmol/l) 137.1 (±7.3) 136.4 (±8.7) 137.4 (±6.4) 0.579
Treatment and complications
Mechanical ventilation 45 (63.4) 16 (59.3) 28 (66.7) 0.532
Use of vasopressors 40 (56.3) 14 (51.9) 26 (61.9) 0.409
Sepsis 19 (26.8) 3 (11.1) 16 (38.1) 0.014
Number of AEDs 5 (4–8) 5 (4–7) 6 (4–9) 0.256
Use of anesthetics 44 (62.0) 16 (59.3) 27 (64.3) 0.674
Induction of burst suppression 29 (40.8) 10 (37.0) 19 (45.2) 0.501
Duration of RSE (days) 10 (4–19) 7 (2–13) 14 (7–24) 0.003
Length of hospital stay (days) 19 (11–36) 17 (10–30) 22 (19–32) 0.331
Length of mechanical ventilation (days) 5 (0–17) 1 (0–15) 9 (0–21) 0.193
AED antiepileptic drug, mRS modified Rankin Scale, SE status epilepticus, NCSE nonconvulsive status epilepticus, RSE refractory status
epilepticus, STESS status epilepticus severity score
Values are n (%), mean (±standard deviation) or median (interquartile range); Parameters significantly (p \ 0.05) associated with long-term
outcome are expressed in bold, parameters showing a statistical trend (p \ 0.1) are expressed in italics

10/71 (14.1 %), progressive symptomatic in 6/71 (8.5 %),
and remote symptomatic in 24/71 (33.8 %) episodes. A
history of seizures was noted in 37/71 (52.1 %) episodes
and the median premorbid mRS was 3 (IQR 1–4; see
Table 1). Anesthetic AEDs were administered in 44/71
(62.0 %) episodes using midazolam in 33/71 (46.5 %),
propofol in 32/71 (45.1 %), sodium thiopental in 19/71
(26.8 %), and ketamine in 5/71 (7.0 %) episodes.
Mortality and functional outcome
The median time from RSE resolution to last available
follow-up was 12 weeks (IQR 6–35). Two (2.8 %)
episodes were lost to follow-up and, therefore, lacked long-
term outcome data. Thirteen of all 71 (18.3 %) episodes
resulted in death during hospital stay (characteristics are
shown in Supplementary Table 1) and 23 patients were
dead at last follow-up. The overall functional outcome is
presented in Fig. 1. The median mRS score was 5 (IQR
4–5) at discharge and 5 (IQR 3–6) at last available follow-
up (Fig. 1a). Good long-term outcome was achieved in
27/69 (39.1 %) episodes (Fig. 1b). Of these, 14 showed
mRS = 0–2 and 13 returned to their premorbid baseline
with mRS [ 2. In sum, 20/71 (28.2 %) episodes returned
to their premorbid baseline during hospital stay and in 5
more cases there was complete recovery after hospital
discharge (characteristics are shown in supplementary

123
 J Neurol

Fig. 1 Outcome after refractory status epilepticus. a Box plot show-
ing median premorbid modified Rankin scale (mRS) scores, mRS
scores at discharge, and at last available follow-up (solid bars). The
box widths represent the interquartile ranges, the whiskers display
maximum and minimum values. Outliers beyond 1.5 interquartile
ranges (IQRs) are depicted as ° or *. b Distribution of premorbid mRS
scores, mRS scores at discharge and at last available follow-up. The
values within the bars are the numbers of episodes in which patients
were assigned the respective mRS scores

Table 2 Multivariable model

Odds ratio 95% confidence interval p value
for prediction of long-term
functional outcome as well as A. Poor long-term outcome
in-hospital mortality
STESS ‡ 3 11.56 1.88–71.04 0.008
Duration of RSE 1.11 1.01–1.22 0.033
Sepsis 10.40 1.24–87.40 0.031
C-reactive protein 1.01 0.99–1.04 0.393
B. In-hospital mortality
STESS C 3 4.99 0.90–27.10 0.066
Mechanical ventilation 7.98 0.61–103.57 0.112
Use of vasopressors 1.87 0.23–14.90 0.831
STESS Status Epilepticus Severity Score, RSE refractory status epilepticus
Parameters significantly (p \ 0.05) associated with poor long-term outcome or in-hospital mortality are
expressed in bold, parameters showing a statistical trend (p \ 0.1) are expressed in italics

Table 2). Poor functional outcome was observed in 47/71
(66.2 %) episodes at hospital discharge and in 42/69
(60.9 %) episodes on the long term. A total of six patients
improved from poor discharge outcome to good long-term
outcome.
Parameters associated with outcome
and performance of STESS in RSE
Table 1 provides an overview of parameters associated
with outcome. The parameter STESS (specifically its
components age and history of seizures) was significantly
associated with poor long-term outcome in univariate
analysis, as were higher C-reactive protein (CRP) levels on
admission, sepsis, and an overall longer RSE duration.
Adjustment upon multivariable analysis revealed that
STESS, RSE duration, and sepsis were independently
related to poor functional long-term outcome, whereas,
only the parameter STESS C 3 was associated with in-
hospital mortality (Table 2). A potentially fatal etiology
was neither linked to poor long-term outcome nor mortal-
ity, but was observed more frequently in cases which did
not achieve full recovery (see Supplementary Table 2). Of
note, none of the RSE treatment-related parameters, as
described previously [5, 22], were significantly linked to
long-term outcome, i.e., neither induction of burst sup-
pression, nor need for vasopressors, nor use of anesthetics.
There was no significant difference in the number of AEDs
applied between patients with good or poor long-term
outcome.
ROC curve analysis confirmed the cut-off dicho-
tomization into STESS C 3 and STESS \ 3 for optimal
discrimination between good and poor outcome [in-hospi-
tal mortality: area under the curve (AUC) = 0.700,
p = 0.050, Youden’s Index (YI) = 0.294, negative pre-
dictive value (NPV) = 0.929, positive predictive value
(PPV) = 0.256; poor long-term outcome: AUC = 0.671,
p = 0.002, YI = 0.368, NPV = 0.607, PPV = 0.756; see
Fig. 2a]. Given its significance upon multivariable analy-
sis, the impact of RSE duration on outcome was

123
J Neurol
Fig. 2 Receiver operating characteristics curves for a Status Epilep-
ticus Severity Score (STESS) and b refractory status epilepticus
(RSE) duration as predictors for in-hospital mortality, poor short- and
long-term outcome. Short-term outcome is the functional status at
discharge, long-term outcome is the functional status at last available
specifically analyzed (Fig. 2b). ROC curve analysis
revealed an RSE duration of 10 days as significant cut-off
point associated with outcome [poor short-term outcome:
AUC = 0.703, p = 0.002, YI = 0.388, NPV = 0.514,
PPV = 0.833; poor long-term outcome: AUC = 0.712,
p = 0.012, YI = 0.310; NPV = 0.545, PPV = 0.750; see
Fig. 2b].
Discussion
The present study analyzed mortality and functional long-
term outcome after RSE, (a) verified the prognostic value
of STESS in RSE, and (b) identified seizure duration as
important parameter related to outcome. Several aspects
deserve attention.
First, our data are in line with previous investigations
which reported mortality rates ranging from 8 to 26 % and
overall poor functional long-term outcome in RSE [3–6,
23, 24]. Confirming recent data, a substantial subset of
patients showed functional improvement over time with
some of them even returning to their premorbid baseline
after hospital discharge [5, 7, 20]. Interestingly, clinical
parameters associated both with good long-term outcome
and with return to baseline were very similar.
Second, STESS was initially created to predict in-hos-
pital mortality in SE [11] and its value in RSE was
uncertain. As demonstrated here, we verified the prognostic
power of STESS in RSE. Our results are in line with those
follow-up. The area under the curve values with the respective 95 %
confidence intervals for each outcome are presented in the legends.
Sensitivity, specificity, and Youden’s Index (YI) are given for the
respective optimal cut-off points (STESS C 3 and RSE duration
C10 days)
of a very recently published study assessing STESS in new-
onset RSE (NORSE) [25]. By using the previously pro-
posed cut-off point, STESS C 3 was independently related
to poor functional long-term outcome in our cohort and we
were able to confirm the high negative predictive value of
STESS C 3 for in-hospital mortality [11, 26]. The potential
of STESS to predict outcome in RSE appears remarkable,
especially given how early in the course of RSE the score is
calculated [11, 27].
Third, although we were unable to confirm the previ-
ously reported association with therapeutic measures like
the use of anesthetics or the induction of burst suppression,
outcome after RSE was clearly influenced by treatment and
complications in our cohort. On the one hand, elevated
CRP levels on admission and development of sepsis during
the hospital stay predisposed poor outcome, a finding
which might reflect that epileptic activity per se leads to
systemic inflammatory reactions [28], or, alternatively,
underscores the impact of certain pro-inflammatory path-
ways on neuronal excitability and therefore seizure
refractoriness [29]. On the other hand, duration of RSE was
strongly associated with overall long-term outcome, while
it did not show significant impact on in-hospital mortality.
The influence of seizure duration on outcome in RSE is
controversially debated and comparing results is difficult
due to heterogeneity in study designs and inconsistent RSE
definitions. In line with our findings, Cooper and coworkers
did not describe a correlation of length of RSE and in-
hospital mortality [6] and explained this observation as a
123

consequence of withdrawal of supportive care and thus
shortened RSE durations in non-survivors. Marchi and
coworkers indirectly confirmed these findings by describ-
ing a longer length of hospital stay in patients surviving SE
treated with therapeutic coma [30]. On the contrary, other
studies reported on significant correlations of longer RSE
duration and poor outcome, but outcome assessment was
limited to functional discharge status in these reports [13,
20, 31]. Follow-up data were presented in the analysis of
Cooper and coworkers [6] and in very recently published
studies on large cohorts of prolonged RSE (PRSE) [32] and
NORSE episodes [25]. In terms of seizure duration and its
impact on long-term functional outcome in RSE of any
etiology, we are the first to report such an association. Of
note, the discrepancy between ours and the results of Lai
and coworkers may be driven by overall longer seizure
durations in PRSE [32].
The main limitation of our study is the single-center
retrospective design. Regarding follow-up, we could not
use data assessed at defined time points but had to rely on
information available in our records. In terms of RSE
duration, not all of our patients received continuous EEG
and therefore RSE duration in some cases may not have
been determined precisely.
One strength of our analysis is the clear definition of
RSE which is in line with previously published work and,
therefore, allows comparability of results [3, 5, 33]. Fur-
thermore, we decided to use a more sophisticated outcome
definition grouping episodes into good outcome even if the
final mRS was [2, as long as RSE did not lead to func-
tional decline. We are aware that such an approach bears
some risk of erroneously classifying episodes as having
good outcome particularly in case of pre-existing severe
disability, as additional injury through RSE may not lead to
an increase in the mRS in such patients. Given the median
premorbid mRS score of 3 in our cohort, our outcome
assessment, however, appears more appropriate than an
approach exclusively using an mRS cut-off for outcome
classification.
In conclusion, STESS and RSE duration are signifi-
cantly associated to outcome after RSE. Thus, rapid initi-
ation of adequate therapy in order to achieve fast
termination of seizures seems crucial. Particularly with
regard to an early use of anesthetic AEDs, it remains to be
established whether STESS can serve as a valuable indi-
cator of how aggressively treatment should be initiated.
Acknowledgments This work was supported by a research grant
(ELAN-13-07-15-1) from the ELAN Fonds University of Erlangen-
Nuremberg, Germany. We would like to thank Inken Martin, MD,
PhD (Victor Chang Cardiac Research Institute, Sydney, Australia) for
critically proofreading the manuscript as well as Olaf Bergmann, MD,
PhD (Karolinska Institute, Stockholm, Sweden) for valuable
discussions.
123
J Neurol
Compliance with ethical standards
Conflicts of interest The authors report no conflicts of interest
relevant to this study.
Ethical standard The study was approved by the local institutional
review board (Ethics Committee Vote: 48_15Bc).
References
1. Hocker S, Wijdicks EF, Rabinstein AA (2013) Refractory status
epilepticus: new insights in presentation, treatment, and outcome.
Neurol Res 35(2):163–168. doi:10.1179/1743132812Y.0000000128
2. Lowenstein DH (2006) The management of refractory status
epilepticus: an update. Epilepsia 47(Suppl 1):35–40. doi:10.1111/
j.1528-1167.2006.00658.x
3. Holtkamp M, Othman J, Buchheim K, Meierkord H (2005) Pre-
dictors and prognosis of refractory status epilepticus treated in a
neurological intensive care unit. J Neurol Neurosurg Psychiatry
76(4):534–539. doi:10.1136/jnnp.2004.041947
4. Rossetti AO, Logroscino G, Bromfield EB (2005) Refractory
status epilepticus: effect of treatment aggressiveness on progno-
sis. Arch Neurol 62(11):1698–1702. doi:10.1001/archneur.62.11.
1698
5. Hocker SE, Britton JW, Mandrekar JN, Wijdicks EF, Rabinstein
AA (2013) Predictors of outcome in refractory status epilepticus.
JAMA Neurol 70(1):72–77. doi:10.1001/jamaneurol.2013.578
6. Cooper AD, Britton JW, Rabinstein AA (2009) Functional and
cognitive outcome in prolonged refractory status epilepticus.
Arch Neurol 66(12):1505–1509. doi:10.1001/archneurol.2009.
273
7. Standley K, Abdulmassih R, Benbadis S (2012) Good outcome is
possible after months of refractory convulsive status epilepticus:
lesson learned. Epilepsia 53(1):e17–e20. doi:10.1111/j.1528-
1167.2011.03327.x
8. Peets AD, Berthiaume LR, Bagshaw SM, Federico P, Doig CJ,
Zygun DA (2005) Prolonged refractory status epilepticus fol-
lowing acute traumatic brain injury: a case report of excellent
neurological recovery. Crit Care 9(6):R725–R728. doi:10.1186/
cc3884
9. Thakur KT, Probasco JC, Hocker SE, Roehl K, Henry B, Kossoff
EH, Kaplan PW, Geocadin RG, Hartman AL, Venkatesan A,
Cervenka MC (2014) Ketogenic diet for adults in super-refractory
status epilepticus. Neurology 82(8):665–670. doi:10.1212/WNL.
0000000000000151
10. Rossetti AO, Milligan TA, Vulliemoz S, Michaelides C, Bertschi
M, Lee JW (2011) A randomized trial for the treatment of
refractory status epilepticus. Neurocrit Care 14(1):4–10. doi:10.
1007/s12028-010-9445-z
11. Rossetti AO, Logroscino G, Milligan TA, Michaelides C, Ruf-
fieux C, Bromfield EB (2008) Status Epilepticus Severity Score
(STESS): a tool to orient early treatment strategy. J Neurol
255(10):1561–1566. doi:10.1007/s00415-008-0989-1
12. Brophy GM, Bell R, Claassen J, Alldredge B, Bleck TP, Glauser
T, Laroche SM, Riviello JJ Jr, Shutter L, Sperling MR, Treiman
DM, Vespa PM (2012) Guidelines for the evaluation and man-
agement of status epilepticus. Neurocrit Care 17(1):3–23. doi:10.
1007/s12028-012-9695-z
13. Sutter R, Marsch S, Fuhr P, Ruegg S (2013) Mortality and
recovery from refractory status epilepticus in the intensive care
unit: a 7-year observational study. Epilepsia 54(3):502–511.
doi:10.1111/epi.12064
J Neurol
14. Lowenstein DH, Bleck T, Macdonald RL (1999) It’s time to revise
the definition of status epilepticus. Epilepsia 40(1):120–122
15. Holtkamp M (2011) Treatment strategies for refractory status
epilepticus. Current opinion in critical care 17(2):94–100. doi:10.
1097/MCC.0b013e328342fab5
16. Pugin D, Foreman B, De Marchis GM, Fernandez A, Schmidt
JM, Czeisler BM, Mayer SA, Agarwal S, Lesch C, Lantigua H,
Claassen J (2014) Is pentobarbital safe and efficacious in the
treatment of super-refractory status epilepticus: a cohort study.
Crit Care 18(3):R103. doi:10.1186/cc13883
17. Sutter R, Marsch S, Fuhr P, Kaplan PW, Ruegg S (2014) Anesthetic
drugs in status epilepticus: risk or rescue? A 6-year cohort study.
Neurology 82(8):656–664. doi:10.1212/WNL.0000000000000009
18. Guidelines for epidemiologic studies on epilepsy (1993) Com-
mission on Epidemiology and Prognosis, International League
Against Epilepsy Epilepsia 34(4):592–596
19. Rossetti AO, Hurwitz S, Logroscino G, Bromfield EB (2006)
Prognosis of status epilepticus: role of aetiology, age, and con-
sciousness impairment at presentation. J Neurol Neurosurg Psy-
chiatry 77(5):611–615. doi:10.1136/jnnp.2005.080887
20. Drislane FW, Blum AS, Lopez MR, Gautam S, Schomer DL
(2009) Duration of refractory status epilepticus and outcome: loss
of prognostic utility after several hours. Epilepsia
50(6):1566–1571. doi:10.1111/j.1528-1167.2008.01993.x
21. Rankin J (1957) Cerebral vascular accidents in patients over the
age of 60 II. Prognosis. Scottish Med J 2(5):200–215
22. Kowalski RG, Ziai WC, Rees RN, Werner JK Jr, Kim G,
Goodwin H, Geocadin RG (2012) Third-line antiepileptic therapy
and outcome in status epilepticus: the impact of vasopressor use
and prolonged mechanical ventilation. Crit Care Med
40(9):2677–2684. doi:10.1097/CCM.0b013e3182591ff1
23. Mayer SA, Claassen J, Lokin J, Mendelsohn F, Dennis LJ,
Fitzsimmons BF (2002) Refractory status epilepticus: frequency,
risk factors, and impact on outcome. Arch Neurol 59(2):205–210
24. Rossetti AO, Reichhart MD, Schaller MD, Despland PA,
Bogousslavsky J (2004) Propofol treatment of refractory status
epilepticus: a study of 31 episodes. Epilepsia 45(7):757–763.
doi:10.1111/j.0013-9580.2004.01904.x
25. Gaspard N, Foreman BP, Alvarez V, Cabrera Kang C, Probasco
JC, Jongeling AC, Meyers E, Espinera A, Haas KF, Schmitt SE,
Gerard EE, Gofton T, Kaplan PW, Lee JW, Legros B, Szaflarski
JP, Westover BM, LaRoche SM, Hirsch LJ (2015) New-onset
refractory status epilepticus: Etiology, clinical features, and out-
come. Neurology. doi:10.1212/WNL.0000000000001940
26. Sutter R, Kaplan PW, Ruegg S (2013) Independent external
validation of the status epilepticus severity score. Crit Care Med
41(12):e475–e479. doi:10.1097/CCM.0b013e31829eca06
27. Rossetti AO, Logroscino G, Bromfield EB (2006) A clinical score
for prognosis of status epilepticus in adults. Neurology
66(11):1736–1738. doi:10.1212/01.wnl.0000223352.71621.97
28. Sutter R, Grize L, Fuhr P, Ruegg S, Marsch S (2013) Acute-phase
proteins and mortality in status epilepticus: a 5-year observational
cohort study. Crit Care Med 41(6):1526–1533. doi:10.1097/
CCM.0b013e318287f2ac
29. Vezzani A, Friedman A, Dingledine RJ (2013) The role of
inflammation in epileptogenesis. Neuropharmacology 69:16–24.
doi:10.1016/j.neuropharm.2012.04.004
30. Marchi NA, Novy J, Faouzi M, Stahli C, Burnand B, Rossetti AO
(2015) Status epilepticus: impact of therapeutic coma on out-
come. Crit Care Med. doi:10.1097/CCM.0000000000000881
31. Claassen J, Hirsch LJ, Emerson RG, Mayer SA (2002) Treatment
of refractory status epilepticus with pentobarbital, propofol, or
midazolam: a systematic review. Epilepsia 43(2):146–153
32. Lai A, Outin HD, Jabot J, Megarbane B, Gaudry S, Coudroy R,
Louis G, Schneider F, Barbarot N, Roch A, Lerolle N, Luis D,
Fourrier F, Renault A, Argaud L, Sharshar T, Gibot S, Bollaert
PE (2015) Functional outcome of prolonged refractory status
epilepticus. Crit Care 19:199. doi:10.1186/s13054-015-0914-9
33. Sutter R, Kaplan PW, Marsch S, Hammel EM, Ruegg S, Ziai WC
(2015) Early predictors of refractory status epilepticus: an inter-
national two-center study. Euro J Neurol Off J Euro Feder Neurol
Soci 22(1):79–85. doi:10.1111/ene.12531
123