Outcome of status epilepticus and the predictive value of the EMSE and STESS scores: A prospective study

Kokoteksti

(1)UEF//eRepository DSpace Rinnakkaistallenteet. https://erepo.uef.fi Terveystieteiden tiedekunta. 2020. Outcome of status epilepticus and the predictive value of the EMSE and STESS scores: A prospective study Sairanen, JJ Elsevier BV Tieteelliset aikakauslehtiartikkelit © British Epilepsy Association CC BY-NC-ND https://creativecommons.org/licenses/by-nc-nd/4.0/ http://dx.doi.org/10.1016/j.seizure.2019.12.026 https://erepo.uef.fi/handle/123456789/8062 Downloaded from University of Eastern Finland's eRepository.

(2) Journal Pre-proof Outcome of status epilepticus and the predictive value of the EMSE and STESS scores: a prospective study Joni J. Sairanen, Anne-Mari Kantanen, Harri T. Hyppölä, Reetta K. Kälviäinen. PII:. S1059-1311(19)30506-0. DOI:. https://doi.org/10.1016/j.seizure.2019.12.026. Reference:. YSEIZ 3631. To appear in:. Seizure: European Journal of Epilepsy. Received Date:. 24 July 2019. Revised Date:. 7 December 2019. Accepted Date:. 31 December 2019. Please cite this article as: Sairanen JJ, Kantanen A-Mari, Hyppölä HT, Kälviäinen RK, Outcome of status epilepticus and the predictive value of the EMSE and STESS scores: a prospective study, Seizure: European Journal of Epilepsy (2020), doi: https://doi.org/10.1016/j.seizure.2019.12.026. This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 Published by Elsevier..

(3) Outcome of status epilepticus and the predictive value of the EMSE and STESS scores: a prospective study Authors. Joni J. Sairanen MDa, Anne-Mari Kantanen MD, PhDa, Harri T. Hyppölä MD, PhDb, Reetta K. Kälviäinen MD, PhDa, c. Affiliations. a. Research institution. Epilepsy Center B3254, Kuopio University Hospital, P.O. Box 100, Kuopio FI 70029, Finland; tel. +35817173005; fax +35817172628. Correspondence. J. Sairanen, Epilepsy Center B3254, Kuopio University Hospital, P.O. Box 100, Kuopio FI 70029, Finland; tel. +358445123111; e-mail: joni.sairanen@fimnet.fi. Number of words. 3998. Number of figures. 2. Number of tables. 3. Funding. The research reported in this manuscript was supported by the Saastamoinen Foundation, the Finnish Epilepsy Research Foundation, and the Maire Taponen Foundation.. Conflicts of interest. JS has received grants from the Finnish Epilepsy Research Foundation and the Maire Taponen Foundation and institutional funding from the Finnish State Research Funding via the Neuro Center, Kuopio University Hospital.. re. -p. ro of. Epilepsy Center, Neuro Center, Kuopio University Hospital, Member of ERN EpiCARE, Kuopio, Finland; bEmergency Department, Kuopio University Hospital, Kuopio, Finland, cInstitute of Clinical Medicine, University of Eastern Finland, Kuopio, Finland. lP. AMK has received a grant from the Finnish Cultural Foundation, a speaker’s honoraria from Orion, Boehringer Ingelheim, MSD, BMS and a travel grant from Sanofi.. na. HH has received a speaker’s honoraria from Orion and Boehringer Ingelheim and an honoraria for the membership of the advisory board from MSD.. ur. RK has received grants from the Academy of Finland and the Saastamoinen Foundation, a speaker’s honoraria from Eisai, UCB, and Orion and an honoraria for the membership of the advisory boards from Eisai, GW Pharmaceuticals, Marinus Pharmaceuticals, Sage Therapeutics, Takeda and UCB.. Jo. Author contributions. JS, AMK, HH and RK designed the study. JS collected and analyzed the study data. HH and RK supervised the data collection and conduct of the study. All authors contributed equally to the data interpretation, literature search, and writing of the manuscript. We confirm that the list of authors includes all persons who fulfill the criteria for authorship. All authors have approved the final manuscript and the listed order of authors..

(4) Highlights -. Status epilepticus had a greater than 40% adverse outcome rate in this cohort Mortality prediction in status epilepticus is possible with simple prognostic tools Functional decline after status epilepticus is common but hard to predict The optimal cutoff point of Status Epilepticus Severity Score was 4 in this study. Abstract Purpose: To assess the short-term outcome of status epilepticus (SE) and test the Epidemiology-based Mortality score in Status Epilepticus (EMSE) and the Status Epilepticus Severity Score (STESS) performance in outcome prediction.. ro of. Methods: Consecutive adults with SE in the Kuopio University Hospital emergency department were recruited between March 23 and December 31, 2015. The one-month outcome was assessed by a combined phone interview and medical record review using the Glasgow Outcome Scale-Extended. The prognostic performance of the EMSE-EAC (EMSE using the combination of etiology, age and comorbidity) and STESS were statistically evaluated.. re. -p. Results: We recorded 151 SE episodes in 137 patients, of whom 47 had a first-time epileptic event (seizure or SE). Of the SE episodes, 9.0% resulted in death, 31.6% in functional decline. For mortality prediction, the AUCs of the EMSE-EAC and STESS were 0.790 (95% CI: 0.633–0.947) and 0.736 (95% CI: 0.559–0.914), respectively. The optimal cutoff points were ≥ 34 for the EMSE-EAC and ≥ 4 for STESS. Negative predictive values for mortality using the EMSE-EAC-34 and STESS-4 were 97.5% and 96.7%, respectively. For functional decline prediction, the EMSE-EAC yielded statistically insignificant results, the STESS performance was poor (AUC = 0.621, 95% CI: 0.519–0.724).. lP. Conclusions: Over 40% of SE patients suffer adverse outcomes. The EMSE-EAC and STESS are useful in shortterm mortality prediction, with a high negative predictive value. The optimized cutoff points for the EMSE-EAC and STESS were ≥ 34 and ≥ 4 for cohort, respectively.. Introduction Background. na. Keywords: Seizure, mortality, GOSE, comorbidity, prognosis, prediction. Jo. ur. Status epilepticus (SE) is an abnormally prolonged seizure state with a high risk of mortality and functional decline. The short-term case fatality rate is 7.6–33% across all seizure types (1). At hospital discharge, 23–58% of survivors have increased need for help in daily living (2–4). Of intensive care patients with refractory SE, less than half recover to baseline function, 20% die and 30% show new functional deficits (5). Aggressive treatment should be aimed at patients who are at a high risk of death or functional deterioration and who may benefit from intensive care. Scoring tools to assess the severity of SE early during its course have been developed to help clinical decision-making. The Status Epilepticus Severity Score (STESS) (6) was the first widely used prognostic tool for SE. Its performance was recently assessed in a systematic review of Wang et al. (7): a pooled area under receiver operating characteristic curve (AUC) of 0.81 was reported for mortality. A sensitivity of 81% and a specificity of 53% were reached with the cutoff point of ≥ 3 (STESS-3). The optimized cutoff point for STESS is still controversial, as some studies have found better prognostic accuracy with the cutoff point of ≥ 4 (STESS-4) (8– 10). The Epidemiology-based Mortality Score in Status Epilepticus (EMSE) was introduced in 2015 (11). With the original combination of four parameters (etiology, age, comorbidity, electroencephalogram), it appears to have a better prognostic performance for mortality than STESS-3 and STESS-4 (12). Few studies have investigated the.

(5) performance of the EMSE and STESS in predicting functional outcome in SE survivors (9,12–14), an equally important outcome variable. Furthermore, the comparison of study findings is difficult because different sets of EMSE parameters and cutoff points have been used in different studies. Study Purpose Our study purposes were (1) to investigate one-month mortality and functional decline after SE and (2) to evaluate the performance of EMSE and STESS in outcome prediction.. Methods Study Design and Setting. ro of. We recruited all consecutive adults (≥ 16 years of age) with SE in the emergency department of Kuopio University Hospital (KUH) between March 23 and December 31, 2015. KUH is a tertiary hospital located in the North Savo region in Eastern Finland. It provides the only around-the-clock emergency neurologic service to a population of 248 000. Recruitment was done by manually exploring the neurologic patient lists. We did not exclude patients who were readmitted for SE. We did not interfere with the treatment of the study patients. A sample size estimate was not made for this study specifically, but we aimed for 150 patients to allow subgroup analyses in our other studies.. -p. We assessed the one-month outcome after hospital discharge by conducting a combined phone interview and medical record review. We opted for phone interviews for several reasons: the patients were scattered across a large geographic area, they often had multiple appointments during the weeks following SE, and we considered it appropriate to reserve personal visits for the treating physician.. re. Access to medical records was authorized by the hospital district in accordance with Finnish legislation. We obtained signed informed consent of the patient or his/her legal representative for the interview. The Committee on Research Ethics of North Savo Hospital District approved the study design. Definition of SE. Measurements. na. lP. The operational SE definition of the Finnish Current Care Guideline (15) was used: a prolonged (> 5 minutes) epileptic seizure, a seizure cluster (≥ 2 discrete seizures) with no interictal recovery, or three or more recurrent tonic-clonic seizures within any 24-hour period. The International League Against Epilepsy (ILAE) definition of SE is different: the time limit for SE depends on the seizure type and there is no separate SE criterion based on recurrent tonic-clonic seizures (16). The SE diagnosis was based on clinical signs. Nonconvulsive SE episodes were verified with electroencephalogram (EEG) using the Salzburg Consensus Criteria (17). Postanoxic SE episodes were excluded. Refractory SE was defined as SE that continued despite first- and second-line drugs.. Jo. ur. An investigator (JS) collected data from electronic medical records and paper forms of emergency medical personnel. The EMSE and STESS were calculated according to the original publications (6,11). For the EMSE, only the “etiology”, “age” and “comorbidity” (EMSE-EAC) parameters were used for practical reasons: the EMSE-EAC is easy to calculate upon patient’s arrival in hospital and it does not require EEG data. Many SE patients are treated and monitored clinically without EEG. In this non-interventional study, we did not make the study patients undergo EEG testing. The SE duration was calculated from the onset of seizure with two exceptions: if the onset of seizure was not witnessed, the duration was calculated from when the patient was discovered. In recurrent tonic-clonic seizures, the duration of SE was calculated from the onset of third seizure—the moment the operational definition of SE was fulfilled according to the Finnish Current Care Guideline (15). We used the Glasgow Outcome Scale-Extended (GOSE) to evaluate outcome at one month after each SE episode. An investigator (JS) conducted phone interviews of eligible patients or their caretakers and reviewed the medical records of all patients to calculate their GOSE scores. A score of 7–8 indicated Good Recovery, 5–6: Moderate Disability, 3–4: Severe Disability, 2: Vegetative State and 1: Death. Analysis.

(6) SPSS 25 was used to analyze data. The outcome variables were death and functional decline in survivors (worsening in the GOSE score, ↓GOSE). Risk factors for adverse outcome were studied by univariate analysis (Fisher’s exact test, Mann-Whitney U-test). The receiver operating characteristic (ROC) curves of the EMSEEAC and STESS were drawn for mortality and ↓GOSE. The AUCs were compared using DeLong’s algorithm. Youden’s index was calculated to find the best cutoff points. Sensitivity, specificity, positive and negative predictive values (PPV, NPV), and accuracy were calculated with both the newly acquired and the originally published cutoff points for the both scores. Cochran’s Q test was used to find differences in score sensitivity, specificity, accuracy, PPV and NPV with different cutoff points. Statistically significant results were confirmed in pairwise analyses using McNemar’s testing. Significance values were adjusted with the Bonferroni correction for multiple testing. Statistical significance was set at p < 0.05.. Results Study Cohort. ro of. We recorded 151 SE episodes in 137 patients during the study period (14 patients were readmitted for SE). Table 1 depicts the patients’ demographics. A first-time epileptic event (seizure or SE) was seen in 47 patients (34.3%), of whom 22 were subject to EEG during their hospital stay. Thirty patients (21.9%) had a Charlson Comorbidity Index score of ≥ 3. Thirty-nine patients (28.5%) were dependent on a caretaker at baseline (GOSE score 3–4, Severe Disability). Cerebrovascular disease (29.9%), dementia of any etiology (22.6%) and diabetes (14.6%) were common comorbidities, as well as psychiatric illnesses with ongoing medical treatment or psychotherapy (34.3%). Alcohol abuse was evident in one-third (33.6%).. -p. Features of the SE Episodes. lP. re. The features of the 151 SE episodes are summarized in Table 1. The course of SE was continuous in 67.5%, a seizure cluster in 23.8%, and a recurrent tonic-clonic seizure in 8.6%. The most common etiology was acute symptomatic (37.7%), most often due to alcohol or antiepileptic drug withdrawal. In contrast, potentially fatal causes, such as drug toxicity, metabolic insult, head trauma, central nervous system infection, and acute cerebrovascular event, were rare. A remote symptomatic etiology (25.2%) was mostly seen in patients with a past cerebrovascular event. Progressive symptomatic episodes (11.3%) were frequently caused by degenerative brain diseases or brain tumors. The etiology was unknown in 21.9% and a defined electroclinical syndrome in 4.0%.. na. Tonic-clonic SE was the most common seizure type (69.5%), followed by nonconvulsive SE without coma (14.6%) and focal motor SE (9.9%). Other SE types were rare. An EEG study was conducted in 44 SE episodes 29.1%). First-line treatment with benzodiazepines was given in 80.1% of the episodes, and a second-line drug was used in 60.3%. Twenty-six (17.2%) SE episodes stopped without pharmacotherapy. There were eleven episodes (7.3%) of refractory SE, of which seven (4.6%) were given intensive care and third-line treatment—four patients were excluded from intensive care because of a poor overall prognosis.. Age. ur. Table 1. Description of the cohort. Mean ± SD (years). Jo. Range (years). Patients (N = 137) 59.5 (±18.1) 16–92. Median (years). 61.0. Age > 65 years. 57 (41.6%). Male gender. 86 (62.8%). Type of living. Independent at home Nursing home Institutional setting* Homeless. 110 (80.3%) 20 (14.6%) 5 (3.6%) 2 (1.5%). First-time epileptic event. 47 (34.3%). Preexisting epilepsy. 68 (49.6%).

(7) Charlson Comorbidity Index score 0. 54 (39.4%). 1–2. 53 (38.7%). ≥3. 30 (21.9%) Seizure episodes (N = 151). SE course Continuous. 102 (67.5%). Seizure cluster. 36 (23.8%). Recurrent tonic-clonic seizure. 13 (8.6%). SE etiology Acute symptomatic. 57 (37.7%). Remote symptomatic. 38 (25.2%). Unknown. 33 (21.9%). Progressive symptomatic. 17 (11.3%). In defined electroclinical syndromes. ro of. 6 (4.0%). SE type before treatment SE with prominent motor phenomena. 122 (80.8%). Tonic-clonic. 105 (69.5%). Focal motor. 15 (9.9%). Myoclonic. 2 (1.3%). SE without prominent motor phenomena. 29 (19.2%) 24 (15.9%). -p. Nonconvulsive SE without coma Nonconvulsive SE with coma. 5 (3.3%). Level of consciousness. 112 (74.2%). re. Stuporous/comatose Confused/somnolent Alert Spontaneous resolution Nonrefractory SE Refractory SE. lP. Treatment response. Length of the hospital stay (h, median [range]) < 30 min 30–60 min > 60 min***. na. SE duration (min, median [range])**. 24 (15.9%) 15 (9.9%) 26 (17.2%) 114 (75.5%) 11 (7.3%) 54 [2–495] 150 [3–4431] 28 (18.7%) 19 (12.7%) 103 (68.7%). Jo. ur. Values are given as N (%) unless otherwise stated. SE, status epilepticus. * Community health center ward or rehabilitation center, or prison ** Missing SE duration data in one episode *** Including the patients who died of refractory SE. Outcome. We successfully obtained the one-month outcome of 133 SE episodes (follow-up rate: 88.1%) (Fig. 1.). The adverse outcome rate was 40.6%: ↓GOSE was registered after 42 episodes (31.6%), and there were 12 case fatalities (9.0%). A new Severe Disability (GOSE score: 3–4), indicating increased dependence on caretakers, was noted after 21 episodes (15.8%). A new Moderate Disability (GOSE score: 5–6), indicating loss of work capacity or reduced participation in social or leisure activities, was registered in 17 (12.8%). Four patients (3.0%) had Lower Good Recovery (GOSE score: 7) because they reported only minor symptoms, such as concentration problems, headaches, tiredness and depression. Altogether, these symptoms were reported after 28 episodes (21.1%)..

(8) Of the twelve case fatalities, four had a refractory SE that was not intensive care treated due to futility. In univariate analysis (Table 2), advanced age, progressive symptomatic etiology, high Charlson Comorbidity Index score and refractory SE were associated with mortality. Refractory SE and lack of preexisting epilepsy were associated with ↓GOSE. All seven intensive care treated patients survived, but five acquired functional deficits. Table 2 Outcome predictors: univariate analysis GOSE unchanged Age (years, mean ± SD) 59.3 (±18.3). ↓GOSE. p. Survival. Mortality. p. 60.3 (±15.3) 1.60 (±1.68). 0.851. 59.6 (±17.3). 74.8 (±11.0). 0.003. na. lP. re. -p. ro of. 0.002 Charlson Comorbidity Index 1.19 (±1.43) 0.203 1.33 (±1.53) 3.83 (±2.98) (score, mean ± SD) First-time epileptic event 0.070 0.344 Yes 22 (53.7%) 19 (46.3%) 41 (87.2%) 6 (12.8%) No 57 (71.3%) 23 (28.8%) 80 (93.0%) 6 (7.0%) 0.037 Preexisting epilepsy 1.000 Yes 48 (73.8%) 17 (26.2%) 65 (91.5%) 6 (8.5%) No 31 (55.4%) 25 (44.6%) 56 (90.3%) 6 (9.7%) 0.032 SE etiology 0.689 Acute symptomatic 27 (65.9%) 14 (34.1%) 41 (93.2%) 3 (6.8%) Progressive symptomatic 8 (66.7%) 4 (33.3%) 12 (70.6%) 5 (29.4%) Remote symptomatic 19 (57.6%) 14 (42.4%) 33 (91.7%) 3 (8.3%) Unknown 25 (71.4%) 10 (28.6%) 35 (97.2%) 1 (2.8%) SE type 0.649 0.458 SE with prominent motor 63 (66.3%) 32 (33.7%) 95 (89.6%) 11 (10.4%) phenomena SE without prominent 16 (61.5%) 10 (38.5%) 26 (96.3%) 1 (3.7%) motor phenomena Level of consciousness 0.075 0.880 Stuporous/comatose 53 (60.9%) 34 (39.1%) 87 (89.7%) 10 (10.3%) Confused/somnolent 14 (66.7%) 7 (33.3%) 21 (95.5%) 1 (4.5%) Alert 12 (92.3%) 1 (7.7%) 13 (92.9%) 1 (7.1%) SE course 1.000 0.071 Continuous 53 (65.4%) 28 (34.6%) 81 (88.0%) 11 (12.0%) Seizure cluster 21 (63.6%) 12 (36.4%) 33 (100.0%) 0 (0.0%) Recurrent tonic-clonic 5 (71.4%) 2 (28.6%) 7 (87.5%) 1 (12.5%) seizure SE duration (min, median) 145 193 0.214 152 197 0.319 0.048 0.009 SE refractoriness Non-refractory SE 77 (67.5%) 37 (32.5%) 114 (93.4%) 8 (6.6%) Refractory SE 2 (28.6%) 5 (71.4%) 7 (63.6%) 4 (36.4%) Values are given as N (%) unless otherwise stated. GOSE, Glasgow Outcome Scale-Extended; SE, status epilepticus.. ur. All SE episodes (151). Jo. Eligible for interview(84). Phone interview. Not reached (7). Successful phone interview (77). Not eligible for interview(67). Review of patient records. No data (18). Dropout (18). Data available (56). GOSE score obtained (133) Fig. 1. Disposition flow chart of the outcome evaluation after SE. SE, status epilepticus; GOSE, Glasgow Outcome Scale-Extended..

(9) AUC Analyses Fig. 2 shows the ROC curves of the EMSE-EAC and STESS in outcome prediction. In mortality prediction, the AUC of the EMSE-EAC was 0.790. The optimal cutoff was ≥ 34 (EMSE-EAC-34: sensitivity 83.3%, specificity 65.3%, Youden’s index 0.486). The AUC of STESS was 0.736. The best STESS cutoff was ≥ 4 (STESS-4: sensitivity 75.0%, specificity 73.6%, Youden’s index 0.486). There was no statistically significant difference between the AUCs of the two scores.. EMSE-EAC: AUC: 0.600, 95% CI: 0.491–0.708 STESS: AUC: 0.621, 95% CI: 0.519–0.724 AUC comparison: p = 0.715. re. EMSE-EAC: AUC: 0.790, 95% CI: 0.633–0.947 STESS: AUC: 0.736, 95% CI: 0.559–0.914 AUC comparison: p = 0.614. -p. ro of. In functional decline prediction, the AUC of STESS was 0.621. The best score performance was achieved with the cutoff point of ≥ 3 (sensitivity 59.5%, specificity 62.0%, Youden’s index 0.215). The AUC of the EMSE-EAC had a 95% confidence interval of 0.491–0.708. Thus, the EMSE-EAC was not a statistically significant predictor of functional decline. Again, the difference between the AUCs of the two scores was statistically insignificant.. lP. Fig. 2. The receiver operating characteristics of the EMSE-EAC and STESS in outcome prediction. EMSE-EAC, Epidemiology-based Mortality score in Status Epilepticus (etiology, age, comorbidity); STESS, Status Epilepticus Severity Score; AUC, area under ROC curve; CI, confidence interval.. Performance Parameters in Mortality Prediction. ur. na. Table 3 displays the performance parameters of the EMSE-EAC and STESS in mortality prediction with different cutoff points. All scores had an equally high sensitivity. The EMSE-EAC-34 and STESS-4 performed best and were not statistically different from each other: the EMSE-EAC-34 had 65.3% specificity and 66.9% accuracy, STESS-4 73.6% specificity and 73.7% accuracy. The NPVs of the EMSE-EAC-34 and STESS-4 for mortality were 97.5% and 96.7%, respectively. Conversely, the PPVs of the two tests were low: 19.2% for the EMSE-EAC34, 22.0% for STESS-4.. Jo. Table 3. Comparison of the EMSE-EAC and STESS performance in mortality prediction with different cutoff points Sensitivity (%) Cochran’s Q-test: p = 0.801 EMSE-EAC-27 83.3 vs. 83.3 EMSE-EAC-34 83.3 vs. 75.0 83.3 vs. 75.0 STESS-3 83.3 vs. 75.0 83.3 vs. 75.0 75.0 vs. 75.0 STESS-4 Specificity (%) Cochran’s Q-test: p < 0.001 EMSE-EAC-27 53.7 vs. 65.3 (p = 0.119) EMSE-EAC-34 53.7 vs. 54.6 (p = 1.000) 65.3 vs. 54.6 (p = 0.184) STESS-3 53.7 vs. 73.6 (p < 0.001) 54.6 vs. 73.6 (p = 0.001) 65.3 vs. 73.6 (p = 0.578) STESS-4 Accuracy (%) Cochran’s Q-test: p < 0.001 EMSE-EAC-27.

(10) 56.4 vs. 66.9 (p = 0.141) EMSE-EAC-34 56.4 vs. 56.4 (p = 1.000) 66.9 vs. 56.4 (p = 0.141) STESS-3 56.4 vs. 73.7 (p = 0.001) 56.4 vs. 73.7 (p = 0.001) 66.9 vs. 73.7 (p = 0.871) STESS-4 The rows are read from left to right. EMSE-EAC, Epidemiology-based Mortality score in Status Epilepticus (etiology, age, comorbidity); STESS, Status Epilepticus Severity Score.. Discussion In this prospective cohort study, the one-month adverse outcome rate of SE was 40.6%. The mortality of 9.0% is among the lowest reported, yet the rate of functional decline was 31.6%. These findings are in agreement with previous studies (1–4). Certain aspects may explain the low mortality: the mean patient age (59.5 years) was relatively low, life-threatening acute causes of SE were rare, only 7.3% of the SE episodes were treatment refractory, and our operational definition of SE allowed us to enroll many patients who received early and successful treatment (18).. ro of. Adverse outcome has been linked to a number of clinical factors (19): advanced age, no previous seizures or preexisting epilepsy, comorbidities, severe etiology, low level of consciousness during SE, and refractory SE. We could link almost all of these factors to either mortality or functional decline (↓GOSE) in our cohort. We chose the GOSE as the outcome measure because it emphasizes daily function without overlooking psychosocial aspects. Originally, the Glasgow Outcome Scale was developed for posttraumatic brain injury, but it has seen use in SE research as well (4,20–21). It is among the few outcome measures that can be administered by phone at one month after hospital discharge.. re. -p. Four patients died of refractory SE after having been excluded from intensive care. The decision was made in the patient’s best interest because the anticipated outcome was either death or survival with severe functional impairment. Most patients with a short life expectancy or limited baseline function do not benefit from aggressive treatment. The goal of intensive care should be recovery to baseline activity. All seven intensive care treated patients in our cohort survived, but only two had a full recovery.. lP. Prognostication of SE with the EMSE-EAC and STESS was the main focus of this study. In mortality prediction, the AUCs of the EMSE-EAC (0.790) and STESS (0.736) indicated a fair score performance. The cutoff points of ≥ 34 for the EMSE-EAC and ≥ 4 for the STESS had the best combined sensitivity and specificity. The two scores had no statistically significant differences in sensitivity, specificity or accuracy. Our results do not support the use of either score to predict functional decline.. na. A high NPV for mortality (> 95%) was observed with both the EMSE-EAC-34 and STESS-4. Thus, a low score from either of the tools is a robust indicator of good outcome, which may be used as a justification to treat conservatively in certain clinical situations. Overly aggressive treatment should be avoided in those SE patients who probably have a good outcome (22). Importantly, the PPVs of the both scores were low (< 25%): a high score does not reliably predict high mortality, and it cannot be used as a justification to withhold treatment.. Jo. ur. There are limitations to this study. Our SE definition differs from the standard ILAE definition, which has to be considered when comparing our results to those of other studies. We did not include patients whose SE started in a hospital ward and who never visited the emergency department. We probably missed some SE patients because they were treated outside the study hospital due to special circumstances, for example, do-not-hospitalize orders. There was data loss during the follow-up; the follow-up rate was 88.1%, and the phone interview was conducted in only 50.1%. Many patients were severely disabled even before SE. We obtained their GOSE score by reviewing medical records only, but we probably missed minor symptoms that could have been registered in the interview. The results from this single-center study may not be generalizable to other populations.. Conclusion. Patients with SE have a high risk of mortality and morbidity. The EMSE-EAC and STESS are useful tools in the prediction of short-term mortality but, unfortunately, not in predicting short-term functional disability after SE. Both scores are easy to calculate and have an excellent NPV for mortality. Our results support the use of cutoff points ≥ 34 for the EMSE-EAC and ≥ 4 for the STESS in predicting mortality. Better prognostic tools for predicting functional outcome should be developed..

(11) Abbreviations AUC: area under receiver operating characteristic curve; EEG: electroencephalogram; EMSE: Epidemiologybased Mortality score in Status Epilepticus; GOSE: Glasgow Outcome Scale-Extended; ILAE: International League Against Epilepsy; KUH: Kuopio University Hospital; NPV: negative predictive value; PPV: positive predictive value; ROC: receiver operating characteristic; SE: status epilepticus; STESS: Status Epilepticus Severity Score Funding The research reported in this manuscript was supported by the Saastamoinen Foundation, the Finnish Epilepsy Research Foundation, and the Maire Taponen Foundation. Acknowledgements We thank biostatistician MSc Tuomas Selander in the Science Service Center of Kuopio University Hospital for assistance with the data analysis and research nurse Pirjo Lavi in the Epilepsy Center of Kuopio University Hospital for her help in recruiting patients. References. 7.. 8. 9. 10. 11. 12. 13. 14.. 15.. 16. 17.. Jo. 18.. ro of. 6.. -p. 5.. re. 4.. lP. 3.. na. 2.. Logroscino G, Hesdorffer DC, Cascino G, Hauser WA, Coeytaux A, Galobardes B, et al. Mortality after a first episode of status epilepticus in the United States and Europe. Epilepsia 2005;46:46–8. Claassen J, Lokin JK, Fitzsimmons BF, Mendelsohn FA, Mayer SA. Predictors of functional disability and mortality after status epilepticus. Neurology 2002;58:139–42. Belluzzo M, Furlanis G, Stragapede L. Predictors of functional disability at hospital discharge after status epilepticus. Epilepsy Res 2015;110:179–82. Hui AC, Joynt GM, Li H, Wong KS. Status epilepticus in Hong Kong Chinese: aetiology, outcome and predictors of death and morbidity. Seizure 2003;12:478–82. Kantanen AM, Reinikainen M, Parviainen I, Kälviäinen R. Long-term outcome of refractory status epilepticus in adults: A retrospective population-based study. Epilepsy Res 2017;133:13–21. Rossetti AO, Logroscino G, Milligan TA, Michaelides C, Ruffieux C, Bromfield EB. Status Epilepticus Severity Score (STESS): a tool to orient early treatment strategy. J Neurol 2008;255:1561–6. Wang P, Guan Y, Lin X, Wang Q, Fu C, Wang X, et al. Diagnostic value of Status Epilepticus Severity Score for survival conditions of patients with status epilepticus: A PRISMA-compliant systematic review and meta-analysis. Epilepsy Behav 2018;82:104–10. Sutter R, Kaplan PW, Rüegg S. Independent external validation of the status epilepticus severity score. Crit Care Med 2013;41:e475–9. Kang BS, Kim DW, Kim KK, Moon HJ, Kim YS, Kim HK, et al. Prediction of mortality and functional outcome from status epilepticus and independent external validation of STESS and EMSE scores. Crit Care 2016;20:25. Reindl C, Knappe RU, Sprügel MI, Sembill JA, Mueller TM, Hamer HM, et al. Comparison of scoring tools for the prediction of in-hospital mortality in status epilepticus. Seizure 2018;56:92–7. Leitinger M, Höller Y, Kalss G, Rohracher A, Novak HF, Höfler J, et al. Epidemiology-based mortality score in status epilepticus (EMSE). Neurocrit Care 2015;22:273–82. Giovannini G, Monti G, Tondelli M, Marudi A, Valzania F, Leitinger M, et al. Mortality, morbidity and refractoriness prediction in status epilepticus: Comparison of STESS and EMSE scores. Seizure 2017;46:31–7. Gao Q, Ou-Yang TP, Sun XL, Yang F, Wu C, Kang T, et al. Prediction of functional outcome in patients with convulsive status epilepticus: the END-IT score. Crit Care 2016;20:46. Rivero Rodríguez D, Scherle Matamoros C, Sam K, DiCapua Sacoto D, Samaniego NM, Pernas Y. Evaluation of STESS, mRSTESS, and EMSE to Predict High Disability and Mortality at Hospital Discharge in Ecuadorian Patients with Status Epilepticus. Neurocrit Care 2018;29:413–8. Kälviäinen R, Bendel S, Jonsson H, Keränen T, Kurola J, Salmi T, et al. Status epilepticus (30.5.2016). Current Care Guidelines. The working group set up by the Finnish Medical Society Duodecim, the Finnish Paediatric Neurology Society and the Finnish Neurological Society. Available online at: www.kaypahoito.fi Trinka E, Cock H, Hesdorffer D, Rossetti AO, Scheffer IE, Shinnar S, et al. A definition and classification of status epilepticus--Report of the ILAE Task Force on Classification of Status Epilepticus. Epilepsia 2015;56:1515–23. Leitinger M, Beniczky S, Rohracher A, Gardella E, Kalss G, Qerama E, et al. Salzburg Consensus Criteria for NonConvulsive Status Epilepticus--approach to clinical application. Epilepsy Behav 2015;49:158–63. Sairanen JJ, Kantanen AM, Hyppölä HT, Kälviäinen RK. Treatment delay in status epilepticus - more effective prehospital symptom recognition warranted. Scand J Trauma Resusc Emerg Med 2019;27:28. Sutter R, Kaplan PW, Rüegg S. Outcome predictors for status epilepticus--what really counts. Nat Rev Neurol 2013;9:525–34. Baysal-Kirac L, Feddersen B, Einhellig M, Rémi J, Noachtar S. Does semiology of status epilepticus have an impact on treatment response and outcome? Epilepsy Behav 2018;83:81–6. Li Y, Tian L, Zeng T, Chen J, Chen L, Zhou D. Clinical features and outcome of super-refractory status epilepticus: a retrospective analysis in West China. Seizure 2014;23:722–7. Marchi NA, Novy J, Faouzi M, Stähli C, Burnand B, Rossetti AO. Status epilepticus: impact of therapeutic coma on outcome. Crit Care Med 2015;43:1003–9.. ur. 1.. 19. 20. 21. 22..

(12)