Coroner and medical examiner documentation of sudden unexplained deaths in epilepsy

Epilepsy Research 68 (2006) 137–143

Coroner and medical examiner documentation of sudden unexplained deaths in epilepsy夽 Paul L. Schraeder a,∗ , Kristen Delin b , Robyn L. McClelland c , Elson L. So d a

Department of Neurology, Drexel University College of Medicine, Hahnemann University Hospital, Philadelphia, PA 19102, United States b Department of Nursing, Thomas Jefferson University, Philadelphia, PA, United States c Division of Biostatistics, Mayo Clinic, Rochester, MN, United States d Department of Neurology, Mayo Clinic, Rochester, MN, United States Received 18 July 2005; received in revised form 6 October 2005; accepted 19 October 2005

Abstract Background: Prevalence data for sudden unexplained death in epilepsy (SUDEP) are hampered by its underuse as a final diagnosis on death certificates in appropriate cases. Few data exist about how coroners (COs) and medical examiners (MEs) in the United States use the diagnosis of SUDEP. Methods: A survey instrument that addressed demographics, professional background, annual cases of epilepsy, seizure history, percentage of post-mortem examinations, cause of death, and use of SUDEP as a diagnosis was sent to all COs and MEs in the United States. Unadjusted comparisons between categorical variables used χ2 tests. A multiple regression model examined the odds of respondents considering SUDEP to be a valid diagnosis. Results: Of 2995 surveys, 80.7% went to COs and 19.3% to MEs. The response rate was 15.9% for COs and 21.8% for MEs. Acknowledgment of SUDEP as a valid entity was greatest among pathologists (83.5%) versus other physicians and nonphysicians (P < .001) and correlated with higher autopsy rates and seeing more cases of epilepsy. In actual practice, SUDEP was not used routinely as a death certificate diagnosis in most cases with no cause of death found at autopsy by any group in the survey regardless of title, educational background, location, autopsy rate, or number of seizure cases seen annually. Conclusions: SUDEP appears to be an underused final diagnosis by COs and MEs throughout the United States. There is a need to educate officials at all levels about this diagnosis in persons who have epilepsy with no other cause of death. © 2005 Elsevier B.V. All rights reserved. Keywords: Coroners; Medical examiners; SUDEP; Survey

Abbreviations: CO(s), coroner(s); GTCS(s), generalized tonic–clonic seizure(s); ME(s), medical examiner(s); PM(s), post-mortem examination(s); SMR(s), standardized mortality ratio(s); SUDEP, sudden unexplained death in epilepsy 夽 This survey was supported by a grant from the Albert Einstein Foundation of Philadelphia and Mayo Foundation for Medical Education and Research. ∗ Corresponding author. 0920-1211/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.eplepsyres.2005.10.004

138

P.L. Schraeder et al. / Epilepsy Research 68 (2006) 137–143

In persons with epilepsy, the mortality rate is 2–3 times greater than that in the general population (Morgan and Kerr, 2002). The spectrum of causes of death that were not related directly to the occurrence of a seizure in persons with epilepsy is similar to that of the general population, with the most common diagnoses being cerebrovascular disease, central nervous system tumor, alcohol-related problems, and congenital neurologic deficits. These causes are associated with standardized mortality ratios (SMRs) that are higher than those seen in the general population (Morgan and Kerr, 2002; Lhatoo et al., 2001). Persons with generalized tonic–clonic seizures (GTCSs) have an increased risk of death (Lhatoo et al., 2001), suggesting that GTCS of whatever cause may be an independent risk factor for death. In some persons with epilepsy, and especially with GTCS, there is no explanation for death, despite careful assessment of history and detailed post-mortem examination (PM) (Sander and Sillanp¨aa¨ , 1998). Death in these persons with epilepsy nearly always occurs suddenly and unexpectedly. The term “sudden unexpected or unexplained death in epilepsy” (SUDEP) is used worldwide to indicate this phenomenon of death. SUDEP is a particularly troubling entity in that most of the victims are young adults (Lathers and Schraeder, 1990), and persons with epilepsy may have a 24-fold greater risk for the occurrence of sudden unexplained death compared to the general population (Ficker et al., 1998). Post-mortem findings in SUDEP have been inconsistent and non-specific, such as pulmonary edema with increased lung weight (Leestma et al., 1989) and myocardial, perivascular, or interstitial fibrosis (Natelson et al., 1998). Thus far, analysis of post-mortem findings reported in the literature has failed to reveal the pathogenesis of SUDEP, although the PM is essential in excluding determinable causes of death. Nonetheless, the pursuit of better understanding of the SUDEP phenomenon and its pathogenesis depends on availability of complete and accurate PMs of SUDEP victims. A relative lack of awareness of SUDEP as an important cause of death in epilepsy has been reported among coroners (COs) and pathologists (Lip and Brodie, 1992; Coyle et al., 1994). A possible explanation for this lack of awareness and potential under reporting of SUDEP may be the myth that seizures per se do not have a fatal consequence (Nashef and Sander, 1996). As a result of a recent

increase in publications on the topic of SUDEP, neurologists have become more aware of this entity. However, SUDEP still engenders much controversy and discussion because the mechanism of death is not known and data on its prevalence and risk factors are inconsistent. Most epidemiologic data come from relatively small selected study populations with limited numbers of subjects and variable post-mortem requirements among the published studies (Walczak et al., 2001). Accurate determination of the national prevalence of SUDEP is hampered by the lack of a mandatory requirement for PM, inaccurate death certificate data, and underuse of SUDEP as a final diagnosis in appropriate cases. We are reporting data from a national survey of COs and medical examiners (MEs) in the United States regarding their practice of PM in determining SUDEP occurrence and their use of SUDEP as a final postmortem diagnosis.

1. Methods We sent a questionnaire to all 2995 COs and MEs in urban, suburban, and rural jurisdictions in the United States. Among the issues addressed in the survey questionnaire were demographics, professional background, annual number of cases of epilepsy, and details collected with respect to seizure history. In addition, the following data were also collected: proportion of autopsies performed in persons with history of seizure disorder, extent of PM, final cause of death, whether SUDEP was acknowledged as a diagnosis, and percentage of unexplained deaths in epilepsy attributed to SUDEP. The Mayo Clinic Survey Research Center developed the survey instrument, which was revised after review by four external experts in SUDEP. Surveys were mailed with preaddressed return envelopes to all COs and MEs. We offered the option of a US$ 10 survey completion payment. Approximately 3 months after the first mailing, a second mailing was made to non-responders. Unadjusted comparisons between categorical variables used χ2 tests. A multiple logistic regression model was used to examine adjusted associations of factors with the odds of considering SUDEP to be a valid diagnosis. This model controlled for factors found to be univariately significant.

P.L. Schraeder et al. / Epilepsy Research 68 (2006) 137–143

2. Results Of the 2995 surveys sent, 80.7% (2418) were sent to COs and 19.3% (577) to MEs. Seventeen percent (510) of the surveys were returned for analysis. The response rate was 15.9% (384) for the COs and 21.8% (126) for MEs. The survey results reflected responses from 66 large cities, 69 suburban regions, 102 small cities, 83 small towns, and 171 rural communities distributed throughout the United States. Some of the respondents covered more than one population center. For our analysis, the large cities, suburban areas, and small cities were coalesced into the urban category and the small towns and rural areas were coalesced into the non-urban category. The differences in the education backgrounds between the responding COs and MEs were significant (χ2 = 197.4, P < .001) (Table 1). The MEs were more likely to have a pathologist or non-pathologist MD background (92% of MEs versus 26.5% of COs). Table 2 addresses whether or not the diagnosis of SUDEP was acknowledged by the respondents as a valid diagnosis. Educational background significantly influenced the rate of recognizing SUDEP as a valid diagnosis (χ2 = 17.7, P < .001). Most respondents who were trained as pathologists (83.5%) recognized SUDEP as a valid post-mortem diagnosis, whereas the percentage of non-pathologist physicians and non-medically trained respondents acknowledging the validity of SUDEP as a post-mortem diagnosis was much lower (63 and 58%, respectively). COs and MEs in urban areas were somewhat more likely to acknowledge SUDEP as a valid diagnosis (χ2 = 4.1, P = .042). Jurisdictions dealing with more cases of epilepsyrelated death were more likely to acknowledge SUDEP as a valid diagnosis (test for trend χ2 = 9.5, P = .002).

139

Also, SUDEP was more likely to be acknowledged in jurisdictions with higher autopsy rates (test for trend χ2 = 14.6, P < .001). There was no significant difference in the recognition of SUDEP as a valid diagnosis as a function of the official title of ME or CO (χ2 = 2.2, P = .136). A multiple logistic regression model adjusted for urban setting, number of epilepsy cases seen per year, professional title, and high percentage of epilepsy cases autopsied found that seeing only some cases of epilepsy per year (Morgan and Kerr, 2002; Lhatoo et al., 2001; Sander and Sillanp¨aa¨ , 1998; Lathers and Schraeder, 1990; Ficker et al., 1998) as opposed to none (OR = 2.14; 95%CI, 1.24–3.67; P = .006) and a high percentage (75–100%) of autopsies (OR = 2.06; 95%CI, 1.23–3.47; P = .004) were significantly associated with the odds of the respondents considering SUDEP as a valid diagnosis (Table 3). Table 4 shows that only a minority of respondents used SUDEP as a final diagnosis in more than half of the cases in which the diagnosis would have been appropriate, i.e., when no identifiable cause of death was found at autopsy. The actual rate of use of SUDEP as a diagnosis in appropriate cases was low regardless of the educational background of the respondent. Although 83.5% of the pathologists had acknowledged that SUDEP is a valid diagnosis if no cause of death were to be found at autopsy (Table 2), only 22.9% of these same pathologists actually used SUDEP in more than half of the cases eligible for the diagnosis. The nature of the community served (urban versus nonurban), the number of seizure cases encountered per year, and the autopsy rate were not associated with the rate of actually using the diagnosis of SUDEP in appropriate cases.

Table 1 Educational background of respondents Education

Pathologist Non-pathologist physician Other Missing ∗

Coroner*

Medical examiner*

No.

%

No.

%

22 80 247 35

5.7 20.8 64.3 9.1

74 42 10 0

58.7 33.3 7.9 0

Differences in educational background between coroners and medical examiners were significant: χ2 = 197.4, P < .001.

P.L. Schraeder et al. / Epilepsy Research 68 (2006) 137–143

140

Table 2 Sudden unexplained death in epilepsy (SUDEP) acknowledged as a valid diagnosis N

Response to this question, no.

Acknowledge SUDEP as valid diagnosis, no. (%)

Educational background of survey respondent Pathologist 96 Non-pathologist physician 122 Other 257 Missing 35

91 90 166 3

76 (83.5) 57 (63.3) 96 (57.8) 3 (100.0)

Community Urban Not urban Missing

76 396 38

68 279 3

52 (76.5) 177 (63.4) 3 (100.0)

Seizure cases/y 0 1–5 6–15 16 or more Missing

198 220 38 34 20

82 198 35 34 1

40 (48.8) 139 (70.2) 27 (77.1) 25 (73.5) 1 (100.0)

Seizure cases autopsied, % 0 1–25 26–75 76–100 Missing

116 81 61 173 79

56 74 55 153 12

31 (55.4) 41 (55.4) 35 (63.6) 119 (77.8) 6 (50.0)

Professional title Medical examiner Coroner

126 384

107 243

77 (72.0) 155 (63.8)

P* <.001

.042

.002

<.001

.136

∗ Excludes missing data. P values for number of seizure cases per year and number of seizure cases autopsied are based on tests for trend across groups. All other P values are based on standard χ2 tests.

Table 3 Adjusted logistic regressions for the odds of considering sudden unexplained death in epilepsy as a valid diagnosis Variable

Odds ratio (95%CI)

P

Urban setting No Yes

1.00 1.35 (0.63–2.90)

.436

Cases seen/y 0 1–5 6–15 16+

1.00 2.13 (1.24–3.67) 2.05 (0.74–5.66) 1.50 (0.52–4.34)

.006 .166 .454

Title Coroner Medical examiner

1.00 1.10 (0.63–1.92)

.751

Autopsied, % <75 75–100

1.00 2.06 (1.23–3.47)

.006

3. Discussion SUDEP once was thought to be a rare phenomenon (Lathers and Schraeder, 1990; Hirsch and Martin, 1971), but when patients and neurologists finally recognized that it is a major cause of death of young adults with epilepsy, interest in studying the phenomenon increased. Overall, persons with idiopathic epilepsy may have an annual risk of SUDEP of 1/1000, whereas those with symptomatic epilepsy may have an annual risk approaching 1/100 (Annegers and Blakley, 1990). Non-compliant males with poorly controlled epilepsy may have a risk as high as 1/50 per year (Leestma et al., 1989; Leestma, 1990). The mechanism of SUDEP is unknown, but it may be an acute, fatal cardiac dysfunction resulting from seizure-related arrhythmia or asystole (Lathers and Schraeder, 1987; Opherk et al., 2002; Tigaran et al.,

P.L. Schraeder et al. / Epilepsy Research 68 (2006) 137–143

141

Table 4 Percentage use of sudden unexplained death in epilepsy (SUDEP) as a cause of death in epilepsy cases with no identifiable cause at autopsy Variable

P

Use of SUDEP, no. (%) <10%

10–50%

>50%

Educational background Pathologist Non-pathologist physician Other

21 (43.8) 10 (29.4) 14 (34.2)

16 (33.3) 16 (47.1) 13 (31.7)

11 (22.9) 8 (23.5) 14 (34.2)

Community Urban Not urban

15 (41.7) 30 (58.8)

10 (27.8) 10 (19.6)

11 (30.6) 11 (21.6)

Seizure cases/y, no. 1–5 6–15 16+

30 (35.7) 8 (42.1) 8 (38.1)

33 (39.3) 4 (21.1) 8 (38.1)

21 (25.0) 7 (36.8) 5 (23.8)

Seizure-related deaths autopsied, % 0 1–25 26–75 >75

1 (20.0) 6 (31.6) 12 (42.9) 27 (37.5)

2 (40.0) 7 (36.8) 11 (39.3) 25 (34.7)

2 (40.0) 6 (31.6) 5 (17.9) 20 (27.8)

.423

.401

.643

.890

2002). Suspected clues for the mechanism of SUDEP include changes in cardiovascular stability and baroreflex sensitivity, even during the interictal state (Hilz et al., 2002; Lathers and Schraeder, 1982; Mameli et al., 1988). Also, seizure activity can be associated with acute pulmonary edema from increased pulmonary vascular pressure and central apnea that result in fatal anoxia (Langan et al., 2000; Johnston et al., 1997; Nashef et al., 1995; So et al., 2000), suggesting that a combination of acute cardiovascular and pulmonary events related to epileptic discharges may result in death. Accurate post-mortem data are important in helping to understand SUDEP. The primary objective of our survey study was to assess the practice of COs and MEs in the United States in documenting SUDEP in death certificates. Coroners are usually officials who are not medically trained, are often employed part time, and serve smaller cities and towns and rural areas. Nonetheless, they usually utilize pathologists to perform most autopsies. In contrast, the majority of MEs are pathologists who are employed full time to serve larger populations. Although occasionally in some urban settings the ME may be called coroner, the professional level is that of an ME. We found a high rate of acknowledgment by COs and MEs of SUDEP as a valid diagnosis. Officials who were trained as pathologists were significantly more

likely to view SUDEP as a valid diagnosis in theoretical cases with no cause of death found at autopsy than were non-pathologist physicians and non-medically trained officials. This difference suggests that knowledge of SUDEP, as a valid diagnosis is a function of educational level. The importance of educational level is also emphasized by our data that indicate officials in urban areas, who are more commonly forensic specialists, are more likely to acknowledge SUDEP as a valid diagnosis than are officials in non-urban jurisdictions. Other variables associated with an increased level of acknowledgment of SUDEP include higher number of cases of epilepsy encountered annually and higher autopsy rates, both of which support the premise that officials in larger population urban settings more readily acknowledge the validity of SUDEP as a diagnosis if no other explanation for the death can be found. A more important finding of our study is that despite the high rate of acknowledgment of SUDEP as a valid diagnosis, both COs and MEs have a low rate of using SUDEP as a final diagnosis in death certificates when the diagnosis is appropriate. Comparison of urban and non-urban settings and title of CO versus ME did not find any difference in the odds of the respondents actually using SUDEP as a final diagnosis. Neither exposure to a larger number of seizure cases nor a higher autopsy rate involving subjects with seizure-

142

P.L. Schraeder et al. / Epilepsy Research 68 (2006) 137–143

related deaths was associated with appropriate use of SUDEP as a death certificate diagnosis. Most unexpectedly, experience in encountering a larger number of cases of epilepsy and educational level of the officials correlated with the lowest rate of actual use of SUDEP as a final diagnosis in cases with no other explanation for the death. Thus, our data indicate an unexplained paradox in that the percentage of appropriate use of the diagnosis of SUDEP is remarkably lower than would have been predicted from the responders who most readily acknowledged intellectually that SUDEP is a valid diagnosis when no cause of death is found. Our survey data support the premise that SUDEP as a final diagnosis is underused in the United States because its use on death certificates in appropriate cases is much lower than the rate of its acknowledgment as a valid diagnostic category. A recent study of all 612 epilepsy-related deaths in England and Wales for the year 1997 found a PM rate of 81.4% (Langan et al., 2002). Forty-four deaths were certified as SUDEP by officials, but on review of the CO data, an additional 292 were found to be probable SUDEP. The authors concluded that SUDEP was largely underreported and should be considered as the most common cause of death in 16- to 50-year-olds with epilepsy in the United Kingdom. A major problem with using the SUDEP diagnosis may be the variability in the way COs and MEs indicate the cause of death of persons with a diagnosis of epilepsy when no explanation is found from the history, PM, or both. In these patients, the cause of death is attributed frequently to entities such as status epilepticus, respiratory failure, fatal seizure, or cardiac arrhythmia. SUDEP is usually one of the least used diagnoses in these cases (Sander and Sillanp¨aa¨ , 1998; Leestma et al., 1989; Leestma, 1990). The lack of national standards in the United States for documenting circumstances at time of death, and when and how PMs should be conducted in cases of epilepsy-related deaths, plus the evident bias against using SUDEP as a death certificate diagnosis imply that the forensic data attributed to epilepsy-related deaths are incomplete. The rate of response to our survey was low despite careful design of the study to enhance the response rate. The development of our survey form at the Mayo Clinic Survey Research Center was based on longstanding experience with surveying persons in different settings and backgrounds, including medical profes-

sionals. We incorporated optimal survey features such as reasonable length of the form and ease of completing the form. The form was reviewed by both internal and external experts (see Acknowledgment). We recognized the need to use additional measures to enhance the rate of survey response from medical professionals, as has been stressed in an editorial (Burmeister, 2003). Thus, we provided a US$ 10 incentive, which respondents accepted only rarely. The reason for the low response rate to our survey is not obvious. Our telephone inquiries with several experts, including MEs, did not disclose a potential explanation for the low response rate. We also sent the survey form again to those who did not respond to the first mailing. However, two studies of medical professionals have shown that late responders do not differ significantly from initial responders, suggesting that non-response bias is not necessarily reduced by an increased response rate (Sobal and Ferentz, 1989; Berk, 1985). The source database of COs and MEs available to our study does not contain information sufficient to determine whether there were substantial differences between responders and non-responders. However, there is no obvious systematic sampling bias that we could detect after careful review of the data. It should be noted that responders to our survey represented a cross-section of communities, age groups, and educational background. Our study represents the best effort to understand the practice of COs and MEs in the United States in documenting SUDEP occurrence. It is unique in showing that SUDEP diagnosis in appropriate cases is underused by COs and MEs despite their awareness and acceptance of the validity of the diagnosis. The results of our study underscore the importance of the need for more accurate forensic data if we are to determine accurately the prevalence and the cause of SUDEP. Insights gained from this study about how COs and MEs approach the diagnosis of cause of death in persons with epilepsy will be useful in planning educational programs for COs and MEs for improving the accuracy of final diagnoses. One of the principal goals of such programs should be to increase awareness that SUDEP is one of the most common causes of death in epilepsy and that it should not be avoided as a final diagnosis. Any educational program, including continuing medical education activities, internet-based information, and written materials directed at COs and MEs, should focus on the accuracy of data on death certifi-

P.L. Schraeder et al. / Epilepsy Research 68 (2006) 137–143

cates. Accurate prevalence and population distribution data would be important steps toward unraveling the mystery of SUDEP.

Acknowledgments The authors thank Drs. Claire Lathers, Ilo Leppik, Jan Leestma, and Braxton Wannamaker for review of the survey instrument and helpful suggestions.

References Annegers, J.F., Blakley, S.A., 1990. Patterns of overall and unexplained death mortality among persons with epilepsy. In: Lathers, C.M., Schraeder, P.L. (Eds.), Epilepsy and Sudden Death. Marcel Dekker, New York, pp. 39–52. Berk, M.L., 1985. Interviewing physicians: the effect of improved response rate. Am. J. Public Health. 75, 1338–1340. Burmeister, L.F., 2003. Principles of successful sample surveys (editorial). Anesthesiology 99, 1251–1252. Coyle, H.P., Baker-Brian, N., Brown, S.W., 1994. Coroners’ autopsy reporting of sudden unexplained death in epilepsy (SUDEP) in the UK. Seizure 3, 247–254. Ficker, D.M., So, E.L., Shen, W.K., Annegers, J.F., O’Brien, P.C., Cascino, G.D., et al., 1998. Population-based study of the incidence of sudden unexplained death in epilepsy. Neurology 51, 1270–1274. Hilz, M.J., Devinsky, O., Doyle, W., Mauerer, A., Dutsch, M., 2002. Decrease of sympathetic cardiovascular modulation after temporal lobe epilepsy surgery. Brain 125, 985–995. Hirsch, C.S., Martin, D.L., 1971. Unexpected death in young epileptics. Neurology 21, 682–690. Johnston, S.C., Siedenberg, R., Min, J.K., Jerome, E.H., Laxer, K.D., 1997. Central apnea and acute cardiac ischemia in a sheep model of epileptic sudden death. Ann. Neurol. 42, 588–594. Langan, Y., Nashef, L., Sander, J.W.A.S., 2000. Sudden unexpected death in epilepsy: a series of witnessed deaths. J. Neurol. Neurosurg. Psychiatry 68, 211–213. Langan, Y., Nashef, L., Sander, J.W.A.S., 2002. Certification of deaths attributable to epilepsy. J. Neurol. Neurosurg. Psychiatry 73, 751–752. Lathers, C.M., Schraeder, P.L., 1982. Autonomic dysfunction in epilepsy: characterization of autonomic cardiac neural discharge associated with pentylenetetrazol-induced epileptogenic activity. Epilepsia 23, 633–647. Lathers, C.M., Schraeder, P.L., 1987. Review of autonomic dysfunction, cardiac arrhythmias, and epileptogenic activity. J. Clin. Pharmacol. 27, 346–356.

143

Lathers, C.M., Schraeder, P.L., 1990. Preface. In: Lathers, C.M., Schraeder, P.L. (Eds.), Epilepsy and Sudden Death. Marcel Dekker, New York, pp. xiii–xvi. Leestma, J.E., Walczak, T., Hughes, J.R., Kalelkar, M.B., Teas, S.S., 1989. A prospective study on sudden unexpected death in epilepsy. Ann. Neurol. 26, 195–203. Leestma, J.E., 1990. Sudden unexpected death associated with seizures: a pathological review. In: Lathers, C.M., Schraeder, P.L. (Eds.), Epilepsy and Sudden Death. Marcel Dekker, New York, pp. 61–88. Lhatoo, S.D., Johnson, A.L., Goodridge, D.M., MacDonald, B.K., Sander, J.W.A.S., Shorvon, S.D., 2001. Mortality in epilepsy in the first 11 to 14 years after diagnosis: multivariate analysis of a long-term, prospective, population-based cohort. Ann. Neurol. 49, 336–344. Lip, G.Y., Brodie, M.J., 1992. Sudden death in epilepsy: an avoidable outcome? J. Roy. Soc. Med. 85, 609–611. Mameli, P., Mameli, O., Tolu, E., Padua, G., Giraudi, D., Caria, M.A., et al., 1988. Neurogenic myocardial arrhythmias in experimental focal epilepsy. Epilepsia 29, 74–82. Morgan, C.L., Kerr, M.P., 2002. Epilepsy and mortality: a record linkage study in a UK population. Epilepsia 43, 1251– 1255. Nashef, L., Fish, D.R., Sander, J.W., Shorvon, S.D., 1995. Incidence of sudden unexpected death in an adult outpatient cohort with epilepsy at a tertiary referral centre. J. Neurol. Neurosurg. Psychiatry 58, 462–464. Nashef, L., Sander, J.W.A.S., 1996. Sudden unexpected deaths in epilepsy: where are we now? Seizure 5, 235–238. Natelson, B.H., Suarez, R.V., Terrence, C.F., Turizo, R., 1998. Patients with epilepsy who die suddenly have cardiac disease. Arch. Neurol. 55, 857–860. Opherk, C., Coromilas, J., Hirsch, L.J., 2002. Heart rate and EKG changes in 102 seizures: analysis of influencing factors. Epilepsy Res. 52, 117–127. Sander, J.W.A.S., Sillanp¨aa¨ , M., 1998. Natural history and prognosis. In: Engel Jr., J., Pedley, T.A. (Eds.), Epilepsy: A Comprehensive Textbook, vol. 1. Lippincott-Raven, Philadelphia, pp. 69– 86. So, E.L., Sam, M.C., Lagerlund, T.L., 2000. Postictal central apnea as a cause of SUDEP: evidence from near-SUDEP incident. Epilepsia 41, 1494–1497. Sobal, J., Ferentz, K.S., 1989. Comparing physicians’ responses to the first and second mailings of a questionnaire. Eval. Health Prof. 12, 329–339. Tigaran, S., Molgaard, H., Dam, M., 2002. Atrio-ventricular block: a possible explanation of sudden unexpected death in epilepsy. Acta Neurol. Scand. 106, 229–233. Walczak, T.S., Leppik, I.E., D’Amelio, M., Rarick, J., So, E., Ahman, P., et al., 2001. Incidence and risk factors in sudden unexpected death in epilepsy: a prospective cohort study. Neurology 56, 519–525.