The influence of risk factors, biomarkers and care settings on SUDEP counseling

Although sudden unexpected death in epilepsy (SUDEP) is the most feared epilepsy outcome, there is a dearth of SUDEP counseling provided by neurologists. This may reflect limited time, as well as the lack of guidance on the timing and structure for counseling. We evaluated records from SUDEP cases to examine frequency of inpatient and outpatient SUDEP counseling, and whether counseling practices were influenced by risk factors and biomarkers, such as post-ictal generalized EEG suppression (PGES). We found a striking lack of SUDEP counseling despite modifiable SUDEP risk factors; counseling was limited to outpatients despite many patients having inpatient visits within a year of SUDEP. PGES was inconsistently documented and was never included in counseling. There is an opportunity to greatly improve SUDEP counseling by utilizing inpatient settings and prompting algorithms incorporating risk factors and biomarkers.


Introduction
Sudden unexpected death in epilepsy (SUDEP) is a tragic epilepsy outcome feared by patients and families, with an incidence of 1 in 1000 per year in children and adults [1,2].The American Academy of Neurology and American Epilepsy Society recommend that neurologists routinely counsel patients about SUDEP risk [1,3].Adults with epilepsy and parents of children with epilepsy also want SUDEP information [4].However, neurologists infrequently discuss SUDEP with patients due to limitations of time and knowledge, as well as anxiety and discomfort about these discussions [4][5][6][7].
Neurologists' hesitance to discuss SUDEP likely reflects lack of guidance on initiating these conversations.The timing, frequency, and approach to counseling are not well-characterized and require patientcentered communication and dynamic risk stratification.Although SUDEP mechanisms are incompletely defined, risk factors include frequent generalized tonic-clonic seizures, nocturnal seizures, and medication non-adherence [1,3].Potential SUDEP biomarkers include interictal heart rate variability, autonomic dysregulation, and post-ictal generalized EEG suppression (PGES) [8][9][10].PGES is electrographic activity <10 uV in all electrodes in a bipolar montage, within 30 s of seizure offset [11].PGES is more common after convulsive seizures, particularly with a prolonged tonic phase, and may drive autonomic dysregulation and heart rate variability [8,12].Applying risk factors and biomarkers to SUDEP counseling is poorly defined and there are few risk assessment tools.There is also no consensus on whether SUDEP conversations should be held in inpatient, outpatient or both settings.
We retrospectively evaluated the records of SUDEP cases to examine frequency of inpatient and outpatient counseling, whether PGES was documented in EEG reports and whether these findings influenced counseling practices.

Materials and methods
We retrospectively identified pediatric and adult definite, probable, and near-SUDEP cases at the University of California, San Francisco, and New York University Langone Medical Center between 2009 and 2015.All patients in our study were managed by epileptologists.We reviewed electronic medical records (EMRs) of patients ages 6 months to 65 years admitted to the epilepsy monitoring unit.Deceased patients were identified through EMR review and submission to the National Death Index.For deceased patients, deaths were considered SUDEP if they had definite epilepsy (two seizures >24 h apart or single seizure with EEG epileptiform abnormalities) and met the criteria for SUDEP [13].Patients were excluded if the epilepsy diagnosis was uncertain, or they did not have an epilepsy monitoring unit (EMU) admission with medical records.
Subject data were de-identified and stored on an encrypted password-protected computer.Subject risk was determined as minimal.The data was collected as part of a multi-site study with a central IRB from the primary site, New York University Langone Medical Center.EEG reports were reviewed to identify terms indicating concern for PGES including "attenuation" "suppression" and "electrodecrement."EEGs were reviewed by an epileptologist to confirm they met PGES criteria [11].For cases where there was no mention of PGES or PGES markers, the post-ictal EEGs were re-reviewed by an epileptologist.We assessed if SUDEP counseling occurred inpatient or outpatient based on chart documentation using the search function for the terms "SUDEP" and "Sudden Unexpected Death in Epilepsy."We examined age at death, gender, race, ethnicity, primary language, and history of medication nonadherence.We qualitatively evaluated documentation to determine the rationale for SUDEP counseling.
Ten participants (36 %) had PGES following ≥ 1 seizure based on independent review.PGES was mentioned in the EEG reports of four patients (14 %).Four additional (14 %) reports mentioned "diffuse attenuation", or "diffuse suppression" confirmed to be PGES on EEG review (Fig. 1).Two reports (7 %) did not comment on PGES, attenuation or suppression, but at least one seizure met EEG criteria for PGES.Twelve participants (43%) did not have seizures captured during EMU admissions.No patient with PGES had documented SUDEP discussions.
The median time between last visit and SUDEP was 5 months (IQR 1.5-9).In comparing the timing of outpatient versus inpatient visits to the time of SUDEP, most (79 %;n = 22) had outpatient visits closer to the time of death.Eleven (39 %) patients had an inpatient or outpatient visit within 1 month of death; SUDEP counseling was not documented at any of these visits.Five patients with PGES (50 %) had a visit within 1 month of SUDEP (Table 2, Supplementary File A).

Discussion
SUDEP counseling is challenging for neurologists and best practices are not well-defined [5][6][7].Our results affirm deficient SUDEP counseling, despite modifiable SUDEP risk factors such as medication nonadherence.The limited counseling that occurred was restricted to outpatient settings.Despite many patients having visits within one month of SUDEP, the topic was not discussed at these visits.PGES was not consistently identified or documented in EEG reports and its presence did not influence prevalence of counseling.Although evidence for the role of PGES in SUDEP pathophysiology is mixed, many studies have identified it as a SUDEP risk factor and most acknowledge its association with other defined SUDEP markers [14,15,[8][9][10][11][12]. Mixed reports on its role partly stem from individual differences in its identification and reporting; improved documentation (potentially facilitated by new automated analysis tools) may help resolve this ambiguity [16].This documentation could then be incorporated into SUDEP risk algorithms in combination with other risk factors to prompt conversations in inpatient and outpatient settings.
As our understanding of SUDEP biomarkers (such as PGES) advances, epileptologists should focus on the identification and documentation of these variables.The presence of SUDEP risk factors should, in turn, motivate counseling about this condition.The role of counseling in modifying SUDEP risk remains unproven, but survey data indicate that patients and families want information [4].Counseling may also impact behaviors such as medication adherence that reduce risk.Clinical checklists can be used to encourage SUDEP discussions.A standardized SUDEP risk screener at each clinic visit or EMU admission could prompt risk reassessment and conversations between physicians and their patients [17].One such electronic health record-based screening algorithm doubled rates of SUDEP counseling.The algorithm divided patients into "high" and "low" risk groups and re-stratified patients at each visit with different messaging based on their categorization [18].Similar reminders could also prompt epileptologists to screen for and document SUDEP biomarkers like PGES in EEG reports.
Many participants had inpatient visits within one month of their death during which SUDEP was not discussed.SUDEP counseling should be considered during inpatient stays.EMU admissions are typically an inflection point in epilepsy care and often uncover changes in SUDEP risk through characterization of nocturnal seizure burden or identification of biomarkers.New risk identification can then serve as a natural segue for SUDEP discussions.Patients and families also desire SUDEP information shortly after a new epilepsy diagnosis, which often occurs in the hospital [3,4].A patient-centered approach is still needed to determine the best counseling setting.Some patients feel overwhelmed with  the extensive information provided at epilepsy diagnosis and may prefer information be divided between multiple visits [4].In these cases, it may be best to defer SUDEP conversations to outpatient.Individual patient assessment is critical to tailor the frequency, timing, and content of counseling for maximum efficacy.Study limitations include generalizability to outpatient epilepsy care, where identification of PGES is less feasible.Our study population was primarily limited to white, non-Hispanic, English-speaking patients; our results may not apply to other groups who have high rates of health care disparities and SUDEP.This may be exacerbated by disparities in EMU admissions, accentuating inequities in care.Finally, SUDEP counseling documentation does not assess counseling that was not documented in the medical records which may make our results an under-representation.

Conclusions
We need to improve neurologist SUDEP counseling practices.Improved identification and documentation of PGES may raise clinician awareness about SUDEP risk, and trigger SUDEP counseling, thereby satisfying an unmet patient need.SUDEP counseling may also be an avenue toward reducing modifiable risk factors.Each case needs an individual, patient-centered approach to determine the ideal counseling setting and timing, but inpatient counseling appears to be underutilized.

Fig. 1 .
Fig. 1.Example of PGES.EEG of a young adult with medically refractory generalized epilepsy.Settings with LFF of 1 Hz, HFF of 70 Hz, notch of 60 Hz, sensitivity at 7 µV/mm and page speed at 30 mm/sec.Seizure semiology of behavioral arrest, left gaze deviation, left arm flexion and right arm extension.Electrographically, the seizure had generalized onset with diffuse theta, followed by 8 s of background attenuation at seizure offset as demonstrated above, meeting criteria for PGES.

Table 2
Summary of SUDEP Counseling Characteristics.