HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

Correspondence to:

ARTICLES: Mohamad Koubeissi and Amer Alshekhlee In-hospital mortality of generalized convulsive status epilepticus: A large US sample Neurology 2007; 69: 886-893 [Abstract] [Full text] [PDF]

Correspondence: Submit a response to this article

Correspondence published:

In-hospital mortality of generalized convulsive status epilepticus: A large US sample

Andrea O. Rossetti, Giancarlo Logroscino   (18 October 2007)

Reply from the author

Mohamad Z. Koubeissi, Amer Alshekhlee   (18 October 2007)

In-hospital mortality of generalized convulsive status epilepticus: A large US sample 18 October 2007
Andrea O. Rossetti, CHUV and University of Lausanne Service de Neurologie, CHUV BH-07, CH-1011-Lausanne, Switzerland, Giancarlo Logroscino Send Correspondence to journal: Re: In-hospital mortality of generalized convulsive status epilepticus: A large US sample andrea.rossetti{at}chuv.ch Andrea O. Rossetti, et al.	We read the article by Koubeissi and Alshekhlee with interest. [1] The overall case-fatality of 3.43% appears low compared to previous studies focusing on the same condition because postanoxic patients (who have a poor outcome) were included. Estimation of disease-specific mortality has a major impact on treatment strategy and health policy. Considering hospital-based cohorts, which should come closer to the population studied by Koubeissi, a Turkish series found a mortality of 21%,[2] and a US study 13% (restricting analysis on GCSE).[3] In a prospective, in-hospital series currently being undertaken in Lausanne, mortality of GCSE is 30%. Although the absolute numbers are much lower than Koubeissi’s cohort and the confidence intervals wide, this difference appears important since these studies excluded postanoxic patients. A possible modifying factor in the study of Koubeissi may be represented by the exclusion of “subtle status”, which often follows GCSE [4] and has a significant mortality. Furthermore, in a Californian population-based study including anoxic subjects, the overall case-fatality was 10.7%, declining to 3.5% if analysis was restricted to patients with GCSE as the principal discharge diagnosis.[5] This suggests that classification and coding may introduce major differences among studies. It would also be interesting to know whether the distribution of teaching/non-teaching hospitals in the sample analyzed in this study [1] reflects the distribution in the US. It is possible that a distribution biased toward non-teaching hospitals may lower mortality. As acknowledged by the authors,[1] listed etiologies and complications are only putative since no causal relationship can be inferred given the study design. In this context, we believe that stating that mechanical ventilation is a predictor of poor outcome in GCSE is misleading. Mechanical ventilation may be needed for therapeutic purposes (i.e., coma induction with an anesthetic agent), or for airway protection. The latter might arise from the GCSE itself or from the medication leading to sedation (e.g., benzodiazepines). Without taking these factors into consideration, mechanical ventilation can just be considered a surrogate for poor outcome with no real predicting value. Real predictors such as age, consciousness impairment, and etiology should be variables that are characteristics of GCSE and not related to the treatment.[3] Finally, we agree with the authors about the importance of adjusting for comorbidities, especially in large cohorts where a large proportion of GCSE occur among the elderly. References 1. Koubeissi M, Alshkhlee A. In-hospital mortality of generaized convulsive statsu epilepticus. A large US sample. Neurology 2007;69:886-893. 2. Sagduyu A, Tarlaci S, Hadiye S. Generalized tonic-clonic status epilepticus : causes, treatment, complications and predictors of case fatality. J Neurol 1998;245:640-646. 3. Rossetti AO, Hurwitz S, Logroscino G, Bromfield EB. Prognosis of status epilepticus: Role of etiology, age and consciousness impairment at presentation. J Neurol Neurosurg Psychiatry 2006;77:611-615. 4. Treiman DM. Electroclinical features of status epilepticus. J Clin Neurophysiol 1995;12:343-362. 5. Wu YW, Shek DW, Garcia PA, Zhao S, Johnston SC. Incidence and mortality of generalized convulsive status epilepticus in California. Neurology 2002;58:1070-1076. Disclosure: The authors report no conflicts of interest.
Reply from the author 18 October 2007
Mohamad Z. Koubeissi, Department of Neurology, Case Western Reserve University 11100 Euclid Avenue Cleveland, OH 44106, Amer Alshekhlee Send Correspondence to journal: Re: Reply from the author mohamad.koubeissi{at}uhhospitals.org Mohamad Z. Koubeissi, et al.	We thank Rossetti et al for their interest in our study. The importance of assessing disease-specific case-fatality prompted our analysis of generalized convulsive status epilepticus (GCSE) mortality in such a large cohort. We agree that the exclusion of subtle status epilepticus (SE) may partially account for the low mortality rate in our study. Status epilepticus is a dynamic disorder where semiological and electrographical changes are seen within the course of a single episode. [4] In the Veterans Affairs (VA) cooperative study of GCSE, the 30-day mortality rate of patients who had overt convulsive activity was 26.8%, compared with 64.9% for those with subtle SE. [6] Patients with subtle SE are likely to be given the ICD-9 code of psychomotor status epilepticus (345.7), and thus excluded from our study. Our cohort included 15,370 cases with the principal diagnosis of GCSE but we included only 11,580 cases in the analysis after data cleansing. If the excluded patients (3,790) had concomitant diagnosis of subtle SE--and using the mortality rate of 64.9% reported by the VA study-- the overall mortality would rise to 18.6%. This is comparable to the mortality reported in the Turkish and US studies. [2,3]. The mortality rate of GCSE as the principal discharge diagnosis in the California series, which used a similar hospital discharge database, is similar to our findings.[5] In our series, mortality rate did not differ between patients who were admitted to teaching hospitals and those who were admitted to non-teaching hospitals. [1] This is shown in the univariate analysis (51% for teaching hospitals and 49% for non-teaching hospitals), and through a separate logistic regression model that included hospital teaching status (odds ratio 1.23, 95% confidence interval 0.9 to 1.5, p = 0.056). We are unaware of any studies that compared the outcome of SE in teaching hospitals to that in non-teaching hospitals. Thus, whether the distribution of teaching and non-teaching hospitals in our study resembles that in the US may be unrelated to the overall mortality. As we mentioned in our discussion, mechanical ventilation may be necessitated by the cause of SE or by SE itself. [1] It is plausible that the outcome of mechanical ventilation varies with the indication, but such information is not available in our database. We agree that mechanical ventilation may be considered a surrogate for poor outcome, and this may be an area of investigation that merits prospective assessment. References 6. Treiman DM, Meyers PD, Walton NY, et al. A comparison of four treatments for generalized convulsive status epilepticus. Veterans Affairs Status Epilepticus Cooperative Study Group. N Engl J Med 1998;339:792-798. Disclosure: The authors report no conflicts of interest.