Abstracts

Rationale:
Psychogenic nonepileptic seizures (PNES) are paroxysmal episodes clinically similar to epileptic seizures but without aberrant brain electrical patterns. PNES is thought to be primarily related to psychiatric distress. PNES patients have high rates of psychiatric diagnoses, and around 75% are female. PNES remains understudied, thus the goal of this study was to characterize the clinical epidemiology of a PNES patient population observed at Vanderbilt University Medical Center (VUMC).
Method:
PNES cases were identified by an automated phenotyping algorithm that incorporated ICD codes, CPT codes, and natural language processing. We performed manual chart review to establish the algorithm positive predictive value. We determined which of 1,653 medical diagnoses in the VUMC Electronic Health Records (VUMC-EHR) co-occurred with PNES diagnosis after accounting for important clinical and demographic covariates. We identified patients with a history of sexual-assault trauma, a known risk factor for PNES, within our EHR and examined the relationship between sexual assault trauma and PNES.
Results:
We identified 3,341 adult PNES patients out of 2,346,808 in the VUMC-EHR (prevalence 0·14%). We replicated previously reported associations with psychiatric disorders including post-traumatic stress disorder (PTSD) and identified novel associations with cerebrovascular disease (OR=1·08, p=2·57E-40). We also discovered a very strong association between PNES and sexual assault trauma (OR=10·26, p=5·36E-146) and found that sexual assault trauma mediated nearly a quarter of the association between female sex and PNES in the VUMC-EHR.
Conclusion:
This study characterized the epidemiology of PNES patients in the VUMC-EHR, confirmed previously reported associations, discovered new associations, and clarified the relationship between female sex, and sexual assault trauma exposure, and PNES.
Funding:
:National Institutes of Health; National Science Foundation