Epidemiology of Functional Seizures Among Adults Treated at a University Hospital

Abstract

Importance: Functional seizures (formerly psychogenic nonepileptic seizures), paroxysmal episodes that are often similar to epileptic seizures in their clinical presentation and display no aberrant brain electrical patterns, are understudied. Patients experience a long diagnostic delay, few treatment modalities, a high rate of comorbidities, and significant stigma due to the lack of knowledge about functional seizures.

Objective: To characterize the clinical epidemiology of a population of patients with functional seizures observed at Vanderbilt University Medical Center (VUMC).

Design, setting, and participants: This case-control study included patients with functional seizures identified in the VUMC electronic health record (VUMC-EHR) system from October 1989 to October 2018. Patients with epilepsy were excluded from the study and all remaining patients in the VUMC medical center system were used as controls. In total, the study included 1431 patients diagnosed with functional seizures, 2251 with epilepsy and functional seizures, 4715 with epilepsy without functional seizures, and 502 200 control patients who received treatment at VUMC for a minimum of a 3 years. Data were analyzed from November 2018 to March 2020.

Exposure: Diagnosis of functional seizures, as identified from the VUMC-EHR system by an automated phenotyping algorithm that incorporated International Classification of Diseases, Ninth Revision (ICD-9) codes, International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10) codes, Current Procedural Terminology codes, and natural language processing.

Main outcomes and measures: Associations of functional seizures with comorbidities and risk factors, measured in odds ratios (ORs).

Results: Of 2 346 808 total patients in the VUMC-EHR aged 18 years or older, 3341 patients with functional seizures were identified (period prevalence, 0.14%), 1062 (74.2%) of whom were women and for which the median (interquartile range) age was 49.3 (39.4-59.9) years. This assessment replicated previously reported associations with psychiatric disorders including posttraumatic stress disorder (PTSD) (OR, 1.22; 95% CI, 1.21-1.24; P < 3.02 × 10-5), anxiety (OR, 1.14; 95% CI, 1.13-1.15; P < 3.02 × 10-5), and depression (OR, 1.14; 95% CI, 1.13-1.15; P < 3.02 × 10-5), and identified novel associations with cerebrovascular disease (OR, 1.08; 95% CI, 1.06-1.09; P < 3.02 × 10-5). An association was found between functional seizures and the known risk factor sexual assault trauma (OR, 10.26; 95% CI, 10.09-10.44; P < 3.02 × 10-5), and sexual assault trauma was found to mediate nearly a quarter of the association between female sex and functional seizures in the VUMC-EHR.

Conclusions and relevance: This case-control study found evidence to support previously reported associations, discovered new associations between functional seizures and PTSD, anxiety, and depression. An association between cerebrovascular disease and functional seizures was also found. Results suggested that sexual trauma may be a mediating factor in the association between female sex and functional seizures.

Figure 1.. Algorithm for Detecting Functional Seizures Cases and Controls Within the VUMC-EHR

All Vanderbilt University Medical Center electronic health records (VUMC-EHR) were initially included, then anyone with generalized or focal epilepsy International Classification of Diseases, Ninth Revision (ICD-9) and International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10) codes were excluded. Patients without convulsion or conversion disorder ICD codes were considered cases, while anyone with both a psychogenic nonepileptic seizures (PNES) keyword in their medical record and the presence of the keyword “EEG” (electroencephalogram) was included as a functional seizures case. CPT indicates Current Procedural Terminology.

Figure 2.. PheWAS of VUMC-EHR Functional Seizures Algorithm-Defined Cases vs Algorithm-Defined Controls and Epilepsy Controls

Results are plotted by category of phenotypes, with each category shown in a different color. Only the top 20 associations are labeled to increase visibility of the graph. Upturned triangles represent positive associations with functional seizures case status, while downturned triangles represent negative associations with functional seizures case status. A horizontal blue line indicates the nominal P value level 0.05, while a horizontal orange line indicates the Bonferroni-corrected P value. PheWAS indicates phenome-wide association study; VUMC-EHR, Vanderbilt University Medical Center electonic health records.