Accepted March 29, 2003.
Incidence and Short-term Prognosis of Status Epilepticus in Adults in Bologna, Italy
Article first published online: 19 JUN 2003
Volume 44, Issue 7, pages 964–968, July 2003
How to Cite
Vignatelli, L., Tonon, C., D'Alessandro, R. and Bologna Group for the Study of Status Epilepticus (2003), Incidence and Short-term Prognosis of Status Epilepticus in Adults in Bologna, Italy. Epilepsia, 44: 964–968. doi: 10.1046/j.1528-1157.2003.63702.x
- Issue published online: 19 JUN 2003
- Article first published online: 19 JUN 2003
- Status epilepticus;
Summary: Purpose: To determine the incidence and the 30-day case fatality of status epilepticus (SE) in the adult resident population of the city of Bologna, Italy.
Methods: Over a 1-year period (March 1, 1999 to February 29, 2000), all patients older than 20 years with SE were included. The case-finding method was based on (a) a prospective surveillance of all public general hospitals in the city by neurologic units, and (b) a review of all discharge codes concerning epilepsy.
Results: The crude and standardized annual incidence rate of SE was 13.1 per 100,000 [95% confidence interval (CI), 9.5–17.5] and 10.7 (95% CI, 7.5–13.8). It was higher in the elderly (older than 60 years) than in young adults (26.2 vs. 5.2) and in women than in men (14.9 vs. 11.0). Acute symptomatic SE accounted for 48%, and a cerebrovascular pathology was the most frequently associated etiologic condition (41%). A history of seizures was reported in 39% of patients. The 30-day case fatality was 39% (33% excluding postanoxic patients).
Conclusions: This study reports the first data on the epidemiology of SE in Italy. The incidence rate found in the population of Bologna is in the same range as that of the other European countries. The 30-day case fatality is higher than all the other population studies (both European and American), despite the broadly similar clinical features of patients. Indirect evidence suggests that some inaccurate patient management could have negatively influenced the outcome of SE.
Status epilepticus (SE) is a medical and neurologic emergency whose epidemiology is in the course of definition. Data based on the population of the United States (1,2) and north-central Europe [French-speaking part of Switzerland (3) and Germany (4)] report annual incidence rates between 41 (1) and eight (3) cases per 100,000 inhabitants in the adult age group. Thirty-day case fatalities in adults are grouped ∼26% (1,5) in the United States and ∼8–9% (3,4) in north-central Europe. The risk and prognosis of SE are functions of age, sex, race, and etiology (5,6), but considering the different rates and case fatality among American and European countries, other factors are probably involved.
Because data from southern Europe are not available, we performed a population survey on the incidence rate and 30-day case fatality of SE in the resident adult population of an Italian city (Bologna), with a 1-year prospective surveillance and retrospective evaluation of ICD-9 discharge codes.
Adult residents (20 years and older) of the city of Bologna, Emilia-Romagna region in central Italy, in the period March 1, 1999 to February 29, 2000, were the population at risk. At December 31, 1998, they numbered 336,876 (155,105 men; 46%).
SE was defined as “a single epileptic seizure of >30-min duration or a series of epileptic seizures during which function is not regained between ictal events in a >30-min period,” according to the Guidelines for Epidemiologic Studies on Epilepsy proposed by the Commission on Epidemiology and Prognosis of the International League Against Epilepsy (7).
Operatively, the diagnosis of epileptic seizure was ascertained in the following cases: (a) when epileptic motor phenomena were observed by a neurologist of the team or clearly described by a direct witness (orally or reported on clinic documentation); (b) when clinical myoclonic phenomena were synchronous with epileptiform EEG activities; or (c) when non–motor seizure-like phenomena were concomitant with epileptiform EEG activities. SE by duration was demonstrated and measured (a) by direct witnesses present from the beginning to end of a >30-min seizure, or (b) by direct witnesses present after the real onset and/or before the real end of a seizure or a group of seizures lasting >30 min. In the second case, for a conservative principle, the beginning coincided with the time of first observation and the end with the time of last observed seizures.
Controversial EEG activity cases were finally evaluated and judged by an EEG expert. Comatose patients with EEG epileptiform activities without motor activity (“electrographic SE”) were not included. History of epilepsy was defined as two or more seizure episodes in a lifetime. Case fatality ratio was defined as death occurring from the initiation of SE to 30 days after the termination of SE.
Etiologies of SE were classified according to the Guidelines for Epidemiologic Studies on Epilepsy (7) as follows: (a) SE occurring in close temporal association with systemic or CNS insult, acute symptomatic (head injury, cerebrovascular accident, CNS infection, CNS tumor, postintracranial surgery, toxic, withdrawal, metabolic, multiple causes, undefined); (b) SE owing to conditions resulting in a static encephalopathy, unprovoked remote symptomatic (head injury, cerebrovascular disease, CNS infection or infestation, pre- and perinatal risk factors, alcohol related, postencephalopathic states); (c) SE owing to progressive CNS disorders, unprovoked progressive symptomatic (tumor, infections, autoimmune, metabolic, degenerative); and (d) SE of no clear etiology, idiopathic (reserved for certain partial or generalized epileptic syndromes with particular clinical and EEG characteristics) and cryptogenic (when no factor associated with increased risk of seizures was identified). Multifactorial denomination was added when more than one of these categories were present.
Seizure types in SE were classified according to the Guidelines for Epidemiologic Studies on Epilepsy (7) as follows: (a) generalized SE (no indication of an anatomic localization and no clinical evidence of focal onset) with the subgroups convulsive (tonic, clonic, or tonic–clonic), nonconvulsive (absence), myoclonic; (b) partial SE (evidence of a clinical partial onset) with the subgroups simple (no consciousness impairment), complex (consciousness impairment, amnesia or confusion during or after the seizure), secondarily generalized; and (c) unclassified SE.
Patient identification and collection
Patients with SE were identified both with (a) a prospective observation and (b) the retrospective evaluation of all hospital ICD-9 discharge codes positive for seizures.
The neurologic units of all the public general hospitals of the city and the Department of Neurological Sciences of the University were involved in the observation of possible SE cases. Bologna has three public hospitals (Bellaria, Maggiore, and S. Orsola-Malpighi), two with emergency rooms, and all with emergency facilities and neurologic services or wards. The Department of Neurological Sciences has only scheduled admissions and houses an outpatients' Epilepsy Centre. SE potential cases were identified by local neurologists by the active surveillance of the patients in neurological wards, neurological counsels, EEG recording, and the weekly survey of emergency wards. In Bologna general hospitals, an on-call neurologist and the EEG facilities are available 12 h/day, 5 days/week. SE confirmation was performed by the coordination center (neuroepidemiology unit, Department of Neurological Sciences) after verification of SE criteria. Patient data (demographic, clinical description of SE, laboratory tests, radiologic investigations, medication history) were collected by the coordination center and were entered into an anonymous electronic database system.
All hospital records of adult residents, discharged in the index period, with ICD-9 codes concerning epilepsy (345.X, all extensions) in all positions, were collected and evaluated by the coordination center. Whole documentation was verified to assure that they met the definition of SE.
For the crude incidence rate of SE, the denominator was the resident adult population of the city of Bologna, and the numerator, the number of prospectively and retrospectively identified patients with SE. The 95% confidence interval (CI) was calculated according to the Poisson distribution (8). Specific crude incidence rates also were calculated by age and sex. Age-adjusted incidence rate was calculated by using the 1991 Italian census population.
Forty patients with incident SE were prospectively collected (mean age, 69.3 years; SD, 16.1 years; range, 20–95 years); 17 were men (mean age, 62.6 years; SD, 15.7 years; range, 20–84 years) and 23, women (mean age, 74.3 years; SD, 14.8 years; range, 28–95 years). Of the 555 subjects with 345.X ICD-9 discharge codes, four more cases of SE were retrospectively included (all women: mean age, 62.5 years; range, 46–69 years). Onset of the SE occurred outside the hospitals in 20 subjects and inside in 24 subjects. In the latter group, the ward of observation was an intensive care unit in 21% of cases, an internal medicine group ward in 54%, a neurologic ward in 17%, and another type of ward in 8%. The ward of admission was an intensive care unit in 30% of the SE outpatients, an internal medicine group ward in 50%, a neurologic ward in 10%, and another type of ward in 10%.
The annual crude incidence rate of SE in adult residents of Bologna was 13.1 per 100.000 (95% CI, 9.5–17.5), 11.0 in men (95% CI, 6.4–17.6), 14.9 in women (95% CI, 9.8–21.6). Table 1 reports incidence rates by age and sex. The age-adjusted annual incidence rate was 10.7 per 100,000 (95% CI, 7.5–13.8), 9.7 in men (95% CI, 5.4–16.0), and 11.5 (95% CI, 7.1–17.6) in women.
|Age (yr)||Population||Cases||Incidence rate (Confidence interval)||Population||Cases||Incidence rate||Population||Cases||Incidence rate|
|20–29||24,371||1||4.1 (0.1–22.8)||22,841||1||4.4 (0.1–24.5)||47,212||2||4.2 (0.5–15.2)|
|40–49||24,358||2||8.2 (1.0–29.6)||25,305||1||4.0 (0.1–22.3)||49,663||3||6.0 (1.2–17.5)|
|50–59||25,515||4||15.7 (4.3–40.2)||28,963||2||6.9 (0.8–24.9)||54,478||6||11.0 (4.0–24.0)|
|60–69||23,372||4||17.1 (4.7–43.8)||29,301||7||23.9 (9.6–49.2)||52,673||11||20.9 (10.4–37.4)|
|70–79||18,893||5||26.5 (8.6–61.7)||29,163||8||27.4 (11.8–54.0)||48,056||13||27.1 (14.4–46.3)|
|>79||7,977||1||12.5 (0.3–69.6)||17,135||8||46.7 (20.1–92.0)||25,112||9||35.8 (16.4–68.0)|
Acute symptomatic SE was found in 34%; unprovoked remote symptomatic, 34%; unprovoked progressive symptomatic, 11%; idiopathic/cryptogenic, 7%; and multifactorial, 14% (an acute symptomatic disorder with a remote symptomatic disorder, four cases; an acute symptomatic disorder with a progressive symptomatic disorder, two cases). In provoked and unprovoked symptomatic cases (41 patients), the following disorders/conditions were present (one or more for each patient): cerebrovascular accidents, 41%; systemic metabolic disorders, 24% (postanoxic encephalopathy, four cases); previous head injury, 10%; CNS degenerative disorders, 10%; alcoholism, 7%; CNS tumors, 5%; pre/perinatal risk factors, 2%; intracranial surgery, 2%; and other factors, 5%; in 20% of cases, an undefined acute condition also was present.
History of epilepsy
A history of seizures was reported by 17 (39%) subjects, all but three symptomatic. Twelve patients were receiving long-term antiepileptic therapy (AEDs; four receiving bitherapy).
Generalized tonic–clonic seizures were found in 9% of cases, absence, in 2%; myoclonic, in 16%; simple partial, in 9%; complex partial, in 16%; partial secondarily generalized, in 41%; unclassified SE was found in 7% (in any case, some form of seizure with no clearly classifiable motor manifestations).
Excluding cases with postanoxic encephalopathy or with unavailable data on total duration (5), SE ended within 2 h in nine (25%) subjects, after 2 and within 24 h in 18 (50%) cases, and after 24 h in nine (25%) subjects. Three more subjects died in a condition during which SE could not ruled out (two at >24 h from the onset; one between 2 and 24 h).
In patients receiving pharmacologic treatment (36), the first intervention was with diazepam (DZP) in 78%, phenytoin (PHT) in 14%, and other drugs in 8% (secondary to diagnostic error). Subsequent medication included DZP (37%), lorazepam (LZP; 3%), phenobarbital (PB; 3%), PHT (37%), and other drugs (20%). Excluding cases with unavailable data at the time of first intervention (10), 15 (57.7%) of 26 were treated within 1 h, four (15.4%) after 1 and within 3 h, and seven (26.9%) after 3 h. Onset of the SE outside the hospitals occurred in 20 outpatients: two had spontaneous resolution, three (17%) of 18 were treated during transfer to hospital. In subjects with reliable data on SE resolution after a specific treatment (27), SE stopped in 11 (40.7%) at the first treatment and in another three (11.1%) at the second treatment.
Thirty-day case fatality was 39% (17 subjects). Excluding the four postanoxic cases (all dead), it was 33%. Among provoked acute and multifactorial cases, it was 62% (53% excluding postanoxic cases); among unprovoked symptomatic cases, it was 20%; no death was recorded among idiopathic/cryptogenic cases. Principal clinical factor distribution in dead and living subjects is shown in Table 2.
|Subjects with SE (44)|
|Dead (17)||Living (27)|
|Mean (SD)||71.2 yr (13.4)||67.1 yr (17.1)|
|Median||69 yr||70 yr|
|Range||52–95 yr||20–91 yr|
|Partial secondarily generalized||47.1%||37.0%|
|NA||3 subjects||1 subject|
In the general adult population of the city of Bologna, we found that (a) the annual crude incidence rate of SE was 13.1 per 100,000; (b) the rate increased with age; (c) acute symptomatic SE is the most common; (d) cerebrovascular disease pathology is the most common etiology; (d) a history of epileptic seizures was reported in about one third of patients; (e) frequency of nonconvulsive SE was not negligible; and (f) 30-day case fatality was very high, mainly in acute symptomatic SE.
The definition and classification of SE in our study are similar to those of previous studies, and as did DeLorenzo et al. (1) and Hesdorffer et al. (2), we included patients with postanoxic encephalopathy. Our method of collection of SE patients differed from that of the other surveys that were either prospective (1,3,4) or retrospective (2). We adopted both a prospective surveillance and a retrospective check of the medical records with the ICD code of epilepsy to minimize the number of possible patients lost, because a neurologist was not on call for about two thirds of the time in a week, and emergency departments were surveyed only weekly in the general hospitals of our territory. Nevertheless, even if we cannot rule out that some patients escaped surveillance, we think that a systematic bias was avoided because the majority of patients (86%) were identified outside a neurologic ward, and the data on the clinical features of SE are consistent with the other surveys. In particular, the finding of the increasing risk of SE with age, the majority of acute symptomatic SE (2,3), the cerebrovascular pathology (both in acute and remote cases) as the main etiologic factor (1–3), and the history of epilepsy in about one third of patients (1,3) are in accordance with the other studies. Moreover, as did Coyetaux et al. (3), we captured a good share (18%) of absence or complex partial cases, greater than that of the Richmond survey (1) (5% of absence and complex partial cases in adults), in which an optimal surveillance system was adopted. This finding is an indirect indicator of good surveillance because nonconvulsive SE is the most difficult to diagnose. Finally, as in the study by Hesdorffer et al. (2), SE ended before 24 h in about three fourths of our patients.
The annual incidence of SE in the adults of our population is between those of the other two north-central European countries, the French-speaking part of Switzerland (8.0 per 100,000) (3) and Germany (15.8 per 100,000) (4). For United States adults, higher annual rates have been reported, similar to European rates in Rochester [17.4 per 100,000 (2)], more than double in Richmond [41 per 100,000 (1)]. The differences among the studies can be largely explained by a different distribution of risk factors due to the variation of the population composition (57% nonwhite people in Richmond, with SE incidence of 57 per 100,000 in nonwhite people and 20 per 100,000 in white people of all ages), but other unknown factors are probably involved.
In contrast with the other studies (2–4), we found a higher rate in women than in men. This finding may be the simple effect of chance: the difference is not statistically significant, as the 95% CIs overlap. Another contrasting finding with respect to other surveys (1–4) was the higher 30-day case fatality, even after correction for postanoxic encephalopathy patients. The ratio in our study was ∼33% with respect to ratios ∼20–25% in the United States and 8–10% in Northern European countries. Even if social and geographic differences can justify these discrepancies, differences in the therapeutic management of SE may better explain this finding. From the observation of Bologna hospitals involved in the survey, we collected some indirect evidence suggesting a probable incorrect management of SE in some phases of its clinical evolution. First, guidelines or protocols for SE management were lacking in all four Bologna hospitals. Second, the specific treatment before hospital admission was applied in only ∼17% of outpatients. Although a direct comparison is not possible because of the different composition of the populations at risk, the specific treatment before hospital admission was administered in ∼59% in the Swiss general population (children included) (3). Third, the choice of the drug for the first-line intervention favored the use of DZP in spite of evidence, available at the time of the survey, of a probable enhanced effectiveness of LZP rather than DZP alone or PHT alone (9,10). Even if direct comparisons are not possible, data from general hospital case collections on generalized convulsive SE in the 1980s, from the United States (11) and the Netherlands (12), reported a greater use of combined therapies (benzodiazepine plus PHT) and a larger response rate at the first therapy (60%) (11) with respect to that recorded in Bologna (40%). Data on intervention in Bologna are more similar to those of a population-based survey in Rochester, MN, U.S.A. (13), reporting data from the 1965 through 1984, in which only monotherapy was given, with a response rate of effectiveness of 31% at the first therapy.
In conclusion, the incidence rate of SE in Italian adults is similar to that of Northern European countries, but in contrast, our population had a very high 30-day case fatality. Differences in management during the early stages of the SE could be crucial for evolution and short-term prognosis. Future descriptive epidemiologic studies will have to take into account the setting and quality of SE care to verify other variables influencing short-term prognosis.
Bologna Group for the Study of Status Epilepticus: Azzimondi G, D'Alessandro R, Tonon C, Vignatelli L (Neuroepidemiology unit, Department of Neurological Sciences); Tinuper P, Baruzzi A (Centro per lo studio e la cura delle epilessie “G.M. Corsino,” Department of Neurological Sciences); Meletti S, Pantieri R, Tassinari CA (Department of Neurosciences, Ospedale Bellaria); Crisci M, Sacquegna T (Neurology Ward, Ospedale Maggiore); Guarino M, Stracciari A, (Neurology unit, Policlinico Sant'Orsola-Malpighi).
Acknowledgment: We thank Daniele Agostini (Dipartimento di Salute Pubblica, Bologna), Lamia Bici, Annamaria Borghi, Francesca Dalpozzo, Andrea Strada, and Anne Collins for revising the English. This work was partly supported by “Ministero dell'Università e della Ricerca Scientifica e Tecnologica ex-60%” 1999/2000 grants.
- 8Calculating confidence intervals for rates and ratios. Neuroepidemiology 1983;2: 257–65..