How common is SUDEP: reconsidering the data
[Part 1: Facts; Thurman DJ]
There are some causes of premature death associated with epilepsy that deserve special attention because they may be preventable. Foremost among these are sudden unexpected death in epilepsy (SUDEP), as well as status epilepticus, and injuries (including drowning) resulting from seizures. SUDEP, especially, has received increased attention in recent years, given that its unexpected occurrence is highly traumatic to the families of those affected.
Published estimates of the frequency of SUDEP have varied widely. In the past, many clinicians—to the extent they were aware of its occurrence—considered the risk too low to warrant routine discussion with their patients. But data accrued during the last 3 decades indicate a substantially greater risk than once thought.
In 2014 my colleagues, Dale Hesdorffer, Jacqueline French, and I set out to compile the best data available from published studies and to estimate more reliably the frequency of SUDEP in populations and its impact on public health (Thurman et al., 2014).
We faced challenges. Trying to measure the true incidence of SUDEP in general populations that include all people with epilepsy meant that we had to look beyond studies from epilepsy clinics, whose enrollees tend to include people of higher than usual risk. We focused instead on population-based studies of SUDEP, of which there were only three (Langan et al., 1998; Leestma et al., 1989, Opeskin & Berkovic 2003) that we found to be of high quality and likely to have found most of the cases occurring in their communities. These studies relied on comprehensive efforts by coroners or medical examiners to investigate all possible cases referred to them from throughout their communities. Without such efforts the results would under-estimate the true occurrence of SUDEP.
From the combined data of these three studies we were able to estimate an annual incidence rate of about 1.2 cases per 1,000 persons with epilepsy. Translating this to the United States (US) population in 2013, we estimated about 2,750 cases. We could translate this to the population of the United Kingdom in 2013, estimating nearly 600 cases.
Four other studies provided more detailed information about how the risk of SUDEP changes with age (Lear-Kaul et al., 2005; Opeskin et al., 2000; Terrence et al., 1975; Zhuo et al., 2012). The combined data from these studies showed there were few reported cases in the first decade of life, that the risk rose in adolescence, peaked in the third and fourth decades, and apparently declined after age 50.
These findings represent the traditional measure of the impact of a cause of death on the public’s health—a measure of the public health burden—simply the reported numbers of deaths each year. Yet this is an incomplete measure that fails to take into account the degree of prematurity of these deaths. There are other ways to consider this. The very premature death of an individual in childhood or young adulthood can be weighed more heavily than the death of a person of advanced age who has reached or exceeded her life expectancy. A method of such weighting now commonly used in public health studies is the years of potential life lost (YPLL) measure, which sums the total years lost (up to a standard age of life expectancy) across all lives lost for a given cause of death. We developed a YPLL estimate, in addition to estimating the number of lives lost to SUDEP. Comparing lives lost to other conditions, we found as expected that the numbers of SUDEP deaths were low in relation to leading neurological diseases such as Alzheimer disease, stroke, or Parkinson disease. But when we compared the years of potential life lost for these conditions, the burden from SUDEP was much greater, second only to stroke. In 2010, we estimated there were 100,510 years of potential life lost in the US due to SUDEP (assuming a life expectancy of 75 years).
Finally, for the total risk of SUDEP over a lifetime (the cumulative incidence), we estimated 8% for a person with epilepsy beginning in early childhood. That risk of course is an average, and for individuals could be substantially higher or lower, depending on factors such as their levels of seizure control, types of seizures, and age of epilepsy onset.
All these estimates may change as better data become available. We believe our estimates to be conservative; the true numbers may be even larger. One reason may be an undercount in medical examiner and coroner reports of SUDEP in persons over 50 years old, among whom autopsies and careful death investigations may be less common.
In our report we concluded: “In any event, it is clear from our analysis that the public health burden of SUDEP is substantial and deserves greater attention from clinicians and the medical research community. The estimated annual number of US deaths from SUDEP—2,750—exceed[s] the corresponding number of deaths attributed to sudden infant death syndrome (SIDS), of which 2,063 were reported in 2010 (CDC; 2014). For more than two decades, SIDS has received a great deal of attention from clinicians, researchers, and public health professionals, who through public education have achieved substantial reductions in its occurrence. SUDEP deserves the same attention” (Thurman et al., 2014).
David J Thurman
Emory University School of Medicine, Atlanta, Georgia, USA
How to cite:
Thurman DJ. How common is SUDEP: reconsidering the data. In: Hanna J, Panelli R, Jeffs T, Chapman D, editors. Continuing the global conversation [online]. SUDEP Action, SUDEP Aware & Epilepsy Australia; 2014 [retrieved day/month/year]. Available from: www.sudepglobalconversation.com.