Tuesday, March 21, 2017

Sudden unexpected death in epilepsy: Measures to reduce risk

Brendan Mclean; Rohit Shankar; Jane Hanna; Caryn Jory; Craig Newman.   Sudden Unexpected Death in Epilepsy: Measures to Reduce Risk.  Pract Neurol. 2017;17(1):13-20.


This review looks at the strategies that may help to reduce the risk of sudden unexpected death in epilepsy beyond that of trying to achieve seizure cessation, which is not possible for up to 30% of patients with epilepsy. These strategies include seizure safety checklists, mobile phone technology, telehealth and various devices currently available or in development. We highlight interventions where there is evidence of benefit, and draw attention for the need both to involve patients with epilepsy in risk reduction and to improve communication with those at risk…

From the article 

Strategies to reduce the risk of SUDEP must involve not only people with epilepsy but also their general practitioners. General practitioners have the most medical contact with people with epilepsy, not just for their epilepsy but also for other conditions that may also influence risk, such as depression and substance misuse. The UK's National Institute for Health and Care Excellence epilepsy guidelines from 2004 to 2012 clearly state that the risk of death in epilepsy, in particular SUDEP, should be discussed as a priority at the time of diagnosis.  Its equivalent in Scotland, the Scottish Intercollegiate Guidelines Network epilepsy guidelines 2015, suggests discussion about SUDEP at an 'appropriate time'. Despite this risk, there is a tacit avoidance of discussing the risk of death both in primary and secondary care, while the patient organisations and SUDEP Action (formerly Epilepsy Bereaved) vociferously support the patients' view that such should be made transparent. This article sets out strategies that may help clinicians to improve person-centred communication when dealing with epilepsy, to reduce this risk.

Discussing SUDEP has been controversial and some countries still advocate a paternalistic approach on a 'need to know basis'.  A recent court judgement in Scotland supported the view that people with epilepsy and their families were entitled to be provided with such information at an early stage. The timing of such a discussion should be tailored to individual needs. Sometimes an early discussion is not appropriate; for example, if there is active psychiatric comorbidity or epilepsy has occurred in the setting of a malignant cerebral tumour. In patients with intellectual disability, the family and carers usually wish to know as early as possible. In our practice, we have this discussion using a structured approach and aided by the SUDEP and seizure safety checklist at diagnosis in people with intellectual disability, and at the first follow-up in those without…

There are numerous studies looking at the risk factors for SUDEP. An in-depth review of the literature identified 18 risk factors, leading to our development of a safety checklist. Of the 18 factors, 11 were potentially modifiable, particularly non-adherence to antiepileptic medications, substance misuse, mood disorders and sleep disruption (see https://www.sudep.org/checklist to register, see training videos and access a copy). There is no absolute risk assigned to each item, nor to the sum of the items, but completing the list allows a discussion around risk modification. Detailed analysis and stratification of the risk factors show that some are more significant than others. Despite initial fears that raising the issues might cause unnecessary alarm, patients tend to receive it positively and using the checklist can modify behaviours…

Device Technology

A recent systemic review into available commercial seizure detection devices showed no suitable robust seizure detection and safety technology though some were clearly promising.[26] Most available devices detect movement and/or physiological changes that occur before or during a seizure such as altered blood oxygen levels, heart rate changes, electrical activity in muscles and changes in galvanic skin resistance. Whether we can call seizure-alert dogs a 'device' is debatable.

Movement Sensors

These comprise a pressure sensor map placed under the mattress or sheets to detect an abnormal movement and absence of movement. While weight and sleep movement adjustments can be made, seizure detection rates are variable, with the most successful devices picking up 89% of tonic–clonic seizures, although one study failed to detect any seizures. Specificity is poor, with frequent false positives, so disrupting sleep of both carers and patients. As with all sensor devices for epilepsy, they also raise issues of individual privacy. Nonetheless, these remain the most popular among parents because of their simplicity.


These detect motion and change in velocity in two or three dimensions. Smartphones are particularly good for this. Sensitivity can be as high as 95%, but again specificity is lower. Speed of detection in one study was a median of 17 s with all detected within 30 s. The use of two accelerometers may improve nocturnal seizure detection.

Physiological Changes

Seizure onset is associated with altered autonomic activity, including decreased skin resistance. When combined with an accelerometer, a galvanic device detected 94% of seizures, but with a significant false positive rate.

Heart rate monitors in one study were 100% sensitive for tonic–clonic convulsions, and almost good for myoclonic seizures, but attempts to refine by adding breathing detectors or electromyographic analysis provided no advantage.

Apnoea devices combined with heart rate monitors are attractive in theory, but studies have not yet shown any benefits.


There are no devices for home use, but when combined with video electroencephalogram (EEG), there was 100% sensitivity within 30 s for tonic–clonic convulsions. Thus, the assessment of a more suitable device is in progress.

Video and Infrared Devices

Video monitoring is feasible, but has not been validated by EEG support. Infrared movement monitors reliably correlate with carer-reported activity, which did not necessarily confirm seizures.[ Using infrared spectroscopy to measure blood oxygen changes failed to detect seizures.

Seizure-alert Dogs

There are numerous anecdotal reports of dogs successfully detecting seizures, but no rigorous studies. Dogs may alert to the seizure itself but not to its onset. Dogs may also react both to non-epileptic seizures and epileptic seizures, and so are not specific. One study reported seizure reduction but our group experienced the tragedy of a patient being killed by her dog that was responding to a seizure.

Antisuffocation Pillows

These are often purchased by families, and are advertised on epilepsy support websites, with one study on carbon dioxide retention properties showing theoretical benefits. There is also the advantage that they are cheap and harmless…

Only long-term studies will determine whether safety checklists, telehealth interventions and mobile technologies have an impact. We have used these approaches for several years, and have seen SUDEP in our region fall in the intellectually disabled community from 4–5/year to nil, and in the non-intellectually disabled community from 6 to 1–2/year. How much of this is from the checklist, and how much from increased awareness among clinicians, the media and people with epilepsy or from improved services in general is difficult to determine. However, these interventions do raise the profile of epilepsy mortality and may contribute to enhanced awareness. Device technology is still in its infancy, and we cannot recommend any single device particularly as none has been shown to prevent SUDEP. The ideal drug—one that suppresses all seizure activity and is free from side effects—may not be developed in our lifetime. Furthermore, clinically based interventions will not capture those individuals at high risk. Self-monitoring by people with epilepsy or carers remains important, although there will still be those whose lifestyles put them at risk. Good communication is essential, but current services are not structured with that in mind. Professionals can only do so much, leaving a service gap between the epilepsy professionals, who see a snapshot of a patient's life, and the day-to-day experiences of people with epilepsy. Empowering people with epilepsy to take responsibility for their condition would do much to bridge that gap.


No comments:

Post a Comment