In-hospital costs in patients with seizures and epilepsy after stroke

Transcription

1 FULL-LENGTH ORIGINAL RESEARCH In-hospital costs in patients with seizures and epilepsy after stroke * Alla Guekht, Maria Mizinova, Andrey Ershov, Denis Guz, Igor Kaimovsky, Paolo Messina, and Ettore Beghi SUMMARY Professor Alla Guekht is director of Moscow Research and Clinical Center for Neuropsychiatry. Objectives: To verify the net effect of seizures after stroke on the use of in-hospital health care resources. Methods: Consecutive patients with first-ever stroke were admitted to the stroke unit of a Moscow hospital and followed prospectively until death or discharge. Each patient experiencing seizures was matched for age, sex, stroke type, National Institutes of Health Stroke Scale score at admission, and stroke risk factors to 2+ patients with no seizures, as controls. Resources consumed included length of hospital stay, admission to the intensive care unit (ICU), diagnostic tests, medical consultations and treatments. Cost estimates were based on the Russian National Health Service perspective. Results: The sample comprised 30 patients with in-hospital seizures and 70 matched controls. Patients dying in hospital were 15 of 30 (50%) versus 4 of 70 (5.7%) (p < 0.001). The overall cost of hospital stay was only slightly (nonsignificantly) higher in patients with seizures, but the cost was significantly higher in patients who died than in patients who were discharged alive. Compared to the controls, patients with seizures spent more intensive care unit (ICU) days and required more computed tomography (CT) scans, x-rays, endoscopies, and specialist consultations, causing higher inhospital costs. Significance: In patients with first-ever stroke, seizures per se do not increase the overall in-hospital costs. However, the higher than expected mortality in patients with seizures is associated with additional hospital costs. KEY WORDS: Seizures, Epilepsy, Stroke, Direct costs, Treatment, Russia. Seizures and epilepsy are a common complication of stroke. Early (acute symptomatic) seizures tend to occur in up to 6% of cases 1 3 and late (unprovoked) seizures in 2 4%, 4 carrying increased morbidity and mortality after stroke. 5 In Russia, cerebrovascular disease was identified as Accepted May 15, 2015; Early View publication June 13, *Moscow Research and Clinical Center for Neuropsychiatry, Moscow, Russian Federation; Department of Neurology, Neurosurgery and Genetics, Russian National Research Medical University, Moscow, Russian Federation; Moscow City Hospital No. 12, Moscow, Russian Federation; and IRCCS Mario Negri Institute for Pharmacological Research, Milan, Italy Address correspondence to Ettore Beghi, Laboratory of Neurological Disorders, IRCCS-Mario Negri Institute for Pharmacological Research, Via Giuseppe La Masa 19, Milan, Italy. ettore.beghi@ marionegri.it Wiley Periodicals, Inc International League Against Epilepsy the etiology of epilepsy in 12.3% of cases with localizationrelated epilepsies, with the highest proportion (15%) in the western areas of the country, including Moscow. 6 Risk factors for seizures and epilepsy include stroke severity, cortical lesions, and type and degree of functional disability. 3,7,8 These factors, along with comorbidities in patients presenting seizures after stroke, may partly explain at least the more ominous prognosis, and consequently the burden of the disease and its costs, in patients with poststroke seizures and epilepsy compared to individuals who do not develop seizures. A disabling stroke is associated with higher direct costs. 9 Epilepsy itself is a substantial socioeconomic burden at different levels in Europe. 10 However, in published reports, the net effect of epilepsy and seizures on the outcome of the disease and resource consumption has not been disentangled 1309

2 1310 A. Guekht et al. Key Points The objective of the study was to verify the net effect of seizures after stroke on the use of in-hospital health care resources. 30 patients with first-ever stroke, admitted to the stroke unit of a Moscow hospital, were followed prospectively until death or discharge. Patients were matched for age, sex, stroke type, NIHSS score at admission, and stroke risk factors to 2+ controls with no seizures. Compared to the controls, patients with seizures spent more ICU days, required more tests and specialist consultations, and had more deaths at discharge. Seizures, with higher than expected mortality, are an independent source of additional hospital costs in patients with first stroke. from the underlying clinical conditions. We therefore carried out an observational study to investigate the independent prognostic role of seizures and epilepsy in patients with stroke and their effects on the consumption of medical resources. Our hypothesis was that having seizures and/or epilepsy involve per se greater use of hospital resources and is consequently a source of additional direct costs in a cohort of patients with stroke after event. Materials and Methods This was a prospective single-center cohort study. Consecutive patients with first-ever stroke were included, characterized by cerebral infarction and/or intracerebral hemorrhage and admitted to the stroke unit of the Moscow City Hospital no. 12 from October 2010 to September Patients with transient ischemic attacks, subarachnoid hemorrhage, cerebral vein thrombosis, and brainstem stroke were excluded. Also excluded were patients with a history of seizures before stroke. As specified by the World Health Organization, stroke was defined as a focal (or global) neurologic impairment of sudden onset lasting more than 24 h (or leading to death), of presumed vascular origin. 11 Stroke subtypes (ischemic and hemorrhagic) were further classified on the basis of computed tomography. The side of the stroke was recorded and the severity was classified according to the National Institutes of Health Stroke Scale (NIHSS). 12 Epileptic seizures were defined according to the International League Against Epilepsy (ILAE). 13 Seizure detection was the task of the physicians and nurses who were present during each patient s hospital stay. Seizures were coded as early (within 7 days of the stroke) or late (occurring after 7 days). Documented stroke risk factors and relevant comorbidities were also identified. These included arterial hypertension, diabetes, ischemic heart disease, atrial fibrillation, and pulmonary and renal diseases. Hypertension was defined as blood pressure >140/90 mm Hg at least twice while in hospital after day 1. Diabetes mellitus was defined as preprandial blood glucose >126 mg/dl on two examinations, postprandial glucose >200 mg/dl, or hemoglobin A1c (HbA1c) >8.5%. History of ischemic heart disease, atrial fibrillation, pulmonary disease, and renal disease was obtained from patients past and present records and from ad hoc treatments. We did not do any power calculations because no specific differences were expected between patients with and without seizures. Enrollment was stopped when a convenience sample of 30 patients with early or late seizures was identified during follow-up. Because our research hypothesis was to calculate the extra costs (if any) attributable to epileptic seizures, we controlled for other sources of cost such as disease type, severity, and risk factors. For each patient with seizures, two or more controls were identified among those without seizures and matched for age (5 years), sex, stroke type, NIHSS score at admission, and stroke risk factors. Cases and controls were followed until discharge or death (whichever came first). For each case and matched controls a number of variables were collected from the medical records. These included the main demographic findings (age and sex), the clinical features at hospital admission (stroke type and site, the NIHSS score at admission, and all relevant comorbidities), and the items indicating the health care resources consumed for the management of stroke and its complications during the acute phase and in-hospital follow-up. These include the length of hospital stay, the stay in the intensive care unit (ICU), diagnostic tests and any other test for the assessment of stroke complications and comorbidities, medical consultations, and therapeutic measures. Cost estimates were based on the Russian National Health Service perspective and its implementation in the City of Moscow, which is based on mandatory health care insurance with reimbursement to the health care providers for services and treatments for the management of all diseases. The amount of reimbursement is specific for each diagnosis and constitutes the medical-economic standard. These costs ( standard costs ) are calculated as the costs of hospital stay (daily cost multiplied by the number of days); drugs were included in these costs. 14,15 In the real clinical situation, management was tailored to the needs of each individual patient and all the necessary additional investigations, consultations, and interventions were done. These additional costs (exceeding the standard) were calculated and evaluated separately. Unit costs of specialist consultations and laboratory and instrumental tests were estimated by applying the tariffs approved for Moscow. The total cost consisted of the costs according to the medical-economic standard and the additional costs.

3 1311 Epilepsy and Hospital Costs of Stroke Descriptive statistics were computed on each demographic features, diagnostic, and therapeutic procedure separately. Differences between cases and controls were assessed using the chi-square test and the Student s t-test, where indicated. Total costs in patients with and without seizures were compared by univariate analyses. Post hoc analysis was also done using a generalized linear model to test the total costs in the two groups after adjusting for vital status (dead or alive), an important source of expenditures. Data were analyzed using the SAS package for PC (9.2 version; SAS Institute, Inc., Cary, NC, U.S.A.). Results During the study, 325 consecutive patients were admitted to the Moscow hospital. On September 30, 2012, 30 patients presented seizures while in hospital. Seizures occurred on day 1 in 21 cases, on days 2 3 in five cases, days 4 6 in one case, and days in three cases. Accordingly, 27 patients had early seizures and three late seizures. There were no cases with status epilepticus. Among patients who had no seizures on that date, 70 fulfilled the matching criteria for controls. Cases and controls were followed for median periods of, respectively, 23 days (range 2 56) and 22 days (range 8 72). The demographic and clinical characteristics of cases and controls are illustrated in Table 1. The age at stroke diagnosis ranged from 35 to 85 years in both groups. There was a slight predominance of males in both groups. Ischemic stroke in the left carotid area was predominant in cases and controls. Arterial hypertension, ischemic heart disease, Table 1. Demographic characteristics of patients in case and control groups Patients with stroke and Patients with stroke without seizure (70) Demographics Age, years Male (%) 16 (53.3) 36 (51.4) 0.92 Stroke type (%) Left ischemic 14 (46.7) 35 (50) 0.86 Right ischemic 11 (36.7) 25 (35.7) 0.95 Hemorrhagic 5 (16.7) 10 (14.3) 0.79 Comorbidities and stroke risk factors (%) Diabetes 3 (10.0) 13 (18.6) 0.35 Hypertension 21 (70.0) 59 (84.3) 0.58 Ischemic heart disease 10 (33.3) 23 (32.9) 0.97 Atrial fibrillation 8 (26.8) 21 (30.0) 0.80 Pulmonary disease 8 (26.8) 20 (28.6) 0.88 Renal disease 5 (16.7) 8 (11.4) 0.53 NIHSS NIHSS, NIH Stroke Scale; SD, standard deviation. atrial fibrillation, pulmonary disease, and diabetes were, in decreasing order, the most common comorbidities in both groups (Table 1). Seizures were all focal with or without secondary generalization. Fifty percent of cases (15/30) died in hospital, compared to 5.7% of controls (4/70) (p < 0.001). Among survivors, NIHSS score at discharge was 3.3 (standard deviation) 1.8 in cases versus in controls, a nonsignificant difference. The overall hospital stay was similar in patients with and without seizures (Table 2). However, patients with seizures spent more days in the ICU and fewer days in the neurology ward. Compared to controls, patients with seizures required more computed tomography scans, x-rays and endoscopies, and specialist consultations. The overall cost of hospital stay was slightly (nonsignificantly) higher in patients with seizures and was mostly accounted for by the ICU costs (Table 3). The cost of additional (above the standard) consultations and, except for ultrasound, the cost of diagnostic tests were significantly higher among cases than controls (Table 4). The total cost was significantly higher in patients who died than in patients who were discharged alive (US$ vs US$; p < ). After adjusting for vital status, patients with and without seizures had fairly similar cost estimates (Table S1). Table 2. Resource consumption by cases and controls out Hospital stay, days Days in neurology ward Days in ICU CT, no <0.01 Consultations (for <0.01 one person, no.) X-ray studies, no <0.01 Blood and urine tests, no Blood biochemistry tests, no. Endoscopies, no ICU, intensive care unit; CT, computed tomography; SD, standard deviation. Table 3. Cost of hospital stay per case in US$ (change, October 2012) out Neurology ward ICU Hospital stay ICU, intensive care unit.

4 1312 A. Guekht et al. Table 4. Cost of additional procedures per case in US$ (change, October 2012) Variable Discussion out CT <0.01 Consultations <0.01 Ultrasound Endoscopies CT, computed tomography; SD, standard deviation. These findings indicate that, overall, the in-hospital cost of managing a stroke complicated by seizures and epilepsy is nonsignificantly higher than the cost of stroke without these complications. However, compared to patients who did not experience seizures while in hospital, patients with seizures met increased ICU costs and additional costs related to consultations and diagnostic tests. In the present study, patients experiencing seizures had a significantly higher death risk while in hospital. The total cost of patients who died was significantly higher than the cost of patients discharged alive. Because our cases and controls were comparable in terms of disease severity and comorbidities at hospital admission, we can speculate that seizures are intrinsically associated with higher death risk. Epileptic seizures in stroke patients are a negative prognostic predictor, as shown by the significantly higher mortality during the acute phase. The large majority of our patients (27/30) had early seizures, which are known to have a poor prognosis in stroke patients. 5 There is a substantial range of mortality data in the literature on stroke and seizures, with agreement on higher rates in patients with seizures. Our in-hospital mortality in patients with seizures was within this range. The figures are in keeping with a Canadian cohort study assessing the impact of seizures on morbidity and mortality after stroke. 16 In that study, the seizures after stroke were associated with greater utilization of health care resources, longer hospital stay, and higher short and long-term mortality. However, stroke severity, a factor associated with the development of seizures, might have accounted for the greater resource utilization. A high death rate in patients with stroke who were experiencing seizures was also found in an Italian study. 17 However, in that study the 30-day case-fatality ratio was 29% versus 14% in patients without seizures. According to Burneo et al., 16 patients with seizures had an increased mortality at 30 days (36.2% vs. 16.8%, p < ); in that study, the NIHSS values were similar to those in our cohort. Arboix et al. 18 reported even higher in-hospital mortality rates for patients with atherothrombotic stroke and very early (48 h) seizures than in those without seizures (70% vs. 19.5%, p < 0.001). Our data differ from other reports that early seizures in patients with stroke did not influence hospital mortality. 19,20 The study populations and methods may explain these differences. In several countries, people with epilepsy use health services more and require assistance for a number of medical disorders more frequently than people without epilepsy. 21,22 However, the temporal association between epilepsy and comorbid conditions was not examined. In the Treatment in Geriatric Epilepsy Research (TIGER) study, a retrospective observational study using Veterans Health Administration data, people older than 65 with new-onset epilepsy, had a threefold risk of psychiatric admission and a fivefold risk of medical admission compared to those without epilepsy. 23 However, that study is hardly comparable to ours because there was no control on the presence of comorbidities and several clinical conditions were more frequent in patients with epilepsy than in nonepileptic individuals. In the United Kingdom, the cost of inpatient care in patients with epilepsy was significantly higher than that of the population as a whole. 24 A substantial fraction of this cost was associated with a primary diagnosis of epilepsy. Our study has some strengths and several limitations. The major strength is the control of variables, other than seizures, thought to influence the consumption of resources and the direct costs, that is, the underlying disease severity and the presence of comorbidities. The other strength is the prospective design, which allowed us to standardize our data collection. The first (major) limitation is the small sample size, which may influence the precision of our estimates. However, selection bias is unlikely because all patients with seizures while in hospital during the study were included. A second limitation is the setting, a public hospital in the Moscow area, which may not be representative of the management of stroke and its complications in other countries. A third limitation is the possibility that the greater resource utilization is due to other (unknown) clinical conditions associated with stroke and/or seizures, which were not identified and controlled at the planning stage. The possibility also exists that some differences are due to chance as we did not adjust for multiple comparisons. Finally, we cannot exclude that some of the patients in the control group might have had late seizures during follow-up. Conclusion Our results confirm that seizures per se do not increase the overall in-hospital costs in patients with first-ever stroke. However, the higher than expected mortality in patients with seizures is associated with additional hospital

5 1313 Epilepsy and Hospital Costs of Stroke costs. These findings need to be confirmed in other populations and health care settings. In addition, longer follow-up is needed to verify whether the extra costs attributable to epilepsy persist after hospital discharge. Acknowledgments and Funding Drs. Elisa Bianchi and Eugene Dmitriev are warmly acknowledged for having actively contributed to the revision of the data included in the revised text. Ethical Publication The study protocol was approved by the ethical review committees of the Russian National Research Moscow University and the Moscow City Hospital No. 12. Disclosure of Conflict of Interest Dr. Beghi serves on the editorial advisory boards of Amyotrophic Lateral Sclerosis, Clinical Neurology & Neurosurgery, and Neuroepidemiology; has received money for board membership from Viropharma and Eisai; has received funding for travel and speaker honoraria from UCB-Pharma, Sanofi-Aventis, GlaxoSmithKline, and for educational presentations from GlaxoSmithKline. Dr. Messina has received funding from Sanofi-Aventis, Eisai, Lombardy Region, and the American ALS Association for the data analysis and data management of randomized controlled trial and observational study protocol. Drs. Guekht, Mizinova, Ershov. Guz, and Kaimovsky have no conflicts of interest to disclose. We confirm that we have read the Journal s position on issues involved in ethical publication and affirm that this report is consistent with these guidelines. References 1. So EL, Annegers JF, Hauser WA, et al. Population-based study of seizure disorders after cerebral infarction. Neurology 1996;46: Bladin CF, Alexandrov AV, Bellavance A, et al. Seizures after stroke: a prospective multicenter study. Arch Neurol 2000;57: Beghi E, D Alessandro R, Beretta S, et al. Incidence and predictors of acute symptomatic seizures after stroke. Neurology 2011;77: Slapø GD, Lossius MI, Gjerstad L. Poststrokeepilepsy: occurrence, predictors and treatment. Expert Rev Neurother 2006;6: Gaitatzis A, Sisodiya SM, Sander JW. The somatic comorbidity of epilepsy: a weighty but often unrecognized burden. Epilepsia 2012;53: Guekht A, Hauser WA, Milchakova L, et al. The epidemiology of epilepsy in the Russian Federation. Epilepsy Res 2010;92: Guekht A, Bornstein NM. Seizures after stroke. Handb Clin Neurol 2012;108: Lancman ME, Golimstok A, Norscini J, et al. Risk factors for developing seizures after a stroke. Epilepsia 1993;34: Mittmann N, Seung SJ, Hill MD, et al. Impact of disability status on ischemic strokecosts in Canada in the first year. Can J Neurol Sci 2012;39: Pugliatti M, Beghi E, Forsgren L, et al. Estimating the cost of epilepsy in Europe: a review with economic modeling. Epilepsia 2007;48: Herman B, Schulte BPM, van Luijk JH, et al. Epidemiology of stroke in Tilberg, the Netherlands: the population-based stroke incidence register, introduction and preliminary results. Stroke 1980;11: Brott T, Adams HP Jr, Olinger CP, et al. Measurement of acute cerebral infarction: a clinical examination scale. Stroke 1989;20: Fisher RS, van Emde Boas W, Blume W, et al. Epileptic seizures and epilepsy: definitions proposed by the International League Against Epilepsy (ILAE) and the International Bureau for Epilepsy (IBE). Epilepsia 2005;46: Federal Law No. 323-ФЗ. About the principles of the healthcare system in the Russian Federation The By-Law of the Government of Moscow No. 661-ПП. About the Territorial Program of the Guarantees of the free medical service to the citizens of the Russian Federation in the City of Moscow Burneo JG, Fang J, Saposnik G; Investigators of the Registry of the Canadian Stroke Network. Impact of seizures on morbidity and mortality after stroke: a Canadian multi-centre cohort study. Eur J Neurol 2010;17: Procaccianti G, Zaniboni A, Rondelli F, et al. Seizures in acute stroke: incidence, risk factors and prognosis. Neuroepidemiology 2012;39: Arboix A, Comes E, Garcıa-Eroles L, et al. Prognostic value of very early seizures for in-hospital mortality in atherothrombotic infarction. Eur Neurol 2003;50: Alberti A, Paciaroni M, Caso V, et al. Early seizures in patients with acute stroke: frequency, predictive factors, and effect on clinical outcome. Vasc Health Risk Manag 2008;4: Reith J, Jergensen HS, Nakayama H, et al. Seizures in acute stroke: predictors and prognostic significance: the Copenhagen Stroke Study. Stroke 1997;28: Currie CJ, Morgan CL, Peters JR, et al. The demand for hospital services for patients with epilepsy. Epilepsia 1998;39: Wiebe S, Bellhouse DR, Fallahay C, et al. Burden of epilepsy: the Ontario Health Survey. Can J Neurol Sci 1999;26: Copeland LA, Ettinger AB, Zeber JE, et al. Psychiatric and medical admissions observed among elderly patients with new-onset epilepsy. BMC Health Serv Res 2011;11: Morgan CL, Kerr MP. Estimated cost of inpatient admissions and outpatient appointments for a population with epilepsy: a record linkage study. Epilepsia 2004;45: Supporting Information Additional Supporting Information may be found in the online version of this article: Table S1. Mean estimated costs in cases and controls adjusting for vital status.