Abstracts

Rationale: Psychogenic non-epileptic seizures (PNES) are common and heterogeneous. Though diagnosis and treatment of PNES have become more sophisticated, and risk factors have been identified to differentiate between epilepsy and PNES, no work has been done to predict clinical outcome of PNES patients. We attempted to prospectively identify type and number of risks predictive of post-diagnosis outcome. Methods: Sixty-eight sequential random admits to the Jefferson Epilepsy Care Unit were enrolled under IRB approval. Diagnosis was by history, exam, imaging, and video-EEG of typical seizures and inter-ictal periods. All S’s underwent neuropsych tests and the “PNES Risk Factor Structured Interview” assessing seizure semiology, psycho-social risks, and “abnormal illness behaviors”. Thirty patients were discharged with PNES diagnosis and 32 with epilepsy. Six subjects were excluded; 4 with uncertain diagnosis, and 2 with both epilepsy and PNES. Follow-up failed with 4 PNES and 2 epilepsy S’s. Gender, age, and IQ of the remaining 25 PNES (7 male) and 26 epilepsy (6 male) S’s were equivalent. At mean discharge month 11.7 (Epilepsy) and 9.8 (PNES) a structured interview of outcome variables including: medical utilization, quality of life (QOL), seizure frequency, and employment status was administered. The PNES S’s were further divided into two groups; a “low risk” group with <7total risks (N=9; M=4.44, SD=1.42), and a “high risk” group with ≥7 (N=16; M=8.50, SD=1.37). Results: PNES S’s endorsed a significantly greater number of risk factors than epilepsy S’s (M=7.04, SD=2.41 vs M= 3.31, SD=2.0; p<.001). Post-discharge, PNES S’s consulted more doctors (M= 2.24, SD= 1.96 vs M=.62, SD= .98, p<.001) than Epilepsy S’s. PNES S’s also had lower scores on a 35 point QOL scale (M=19.4, SD=7.9 vs. M=25.7, SD= 8.0, p=.006). At follow-up, S’s in the high and low risk groups did not differ significantly in number of seizures in last 30 days (M=8.06, SD=24.6 vs M=18.1, SD=49.5), doctors seen (M=2.19, SD=1.68 vs M=2.33, SD=2.50), hospital visits (M=.44, SD=.63 vs. M=.22, SD=.67), current AEDs (M=.44 for both groups, SD=.63 vs 0.53), months working full time (M=.75, SD=1.34 vs M=1.78, SD=2.95), or QOL scores (M=17.9, SD=7.5 vs M=22.1, SD=8.3). Finally, the total number of PNES risks was uncorrelated with any outcome variables. Conclusions: The diagnosis of PNES was well predicted by a higher number of risk factors, and PNES patients demonstrated worse outcome in measured domains than even the severely ill epilepsy patients admitted for monitoring. However, a higher number of risks did not predict worse outcome. Personal, historical, and semiologic PNES risk factors may be regarded as vulnerabilities for the emergence of the abnormal illness behavior constituting PNES, but once the behavior emerges its persistence may be determined by patterns of social response to it more than to the vulnerabilities which originally motivated it. The study is weakened by its reliance on patients’ subjective reports of seizure frequency.