Abstracts

Rationale: Previous research has suggested that fibromyalgia (FM) may be an effective clinical marker for the diagnosis of psychogenic non-epileptic seizures (PNES) (Benbadis, 20005). However, this research was focused on a small patient cohort in a tertiary care epilepsy surgery referral center (Benbadis, 2005). The purpose of our case-matched study was to determine how frequently a historical diagnosis of FM was associated with different paroxysmal neurological disorders, and explore the utility of FM as a predictor for the diagnosis of PNES within a mixed population of general neurology patients at a developing epilepsy clinic. Methods: The billing diagnosis codes of 1,730 consecutive, non-selected patient encounters were reviewed over a three-year period for an epileptologist in a neurology clinic to identify all patients with historical diagnoses of FM. The frequency with which epileptic seizures (ES), PNES, and physiological non-epileptic events (PhysNEE) were comorbid with FM was assessed. Age and gender case-matched controls were used for a between-group comparison. Wilcoxon tests were used to analyze interval data, and Chi-squared tests were used to analyze categorical data (p < 0.05). Results: FM was retrospectively identified in 95/1,730 (5.5%) patients in this cohort. Females represented 95% of the FM sample (age: 53 years; 95% CI: 57, 51). Forty-three percent of those with FM had a non-paroxysmal, neurological primary clinical diagnosis, most commonly chronic pain. Paroxysmal events were present in 57% of FM patients and 54% of case-matched controls. Among patients with FM and paroxysmal disorders, 11% had ES, 74% had PNES, and 15% had PhysNEE, compared to case-matched controls with 37% ES, 51% PNES, and 12% PhysNEE (p = 0.009) (Figure 1). Comparison between the FM patients and case-matched controls revealed that a historical diagnosis of FM had a 61% sensitivity, 64% specificity, 74% positive predictive value, and 49% negative predictive value for the diagnosis of PNES. No population differences were present between patients with FM and their corresponding case-matched controls with respect to age, race, marital status, or reason for referral (Figure 2). Conclusions: A historical diagnosis of FM was shown to be a predictor for the diagnosis of PNES in patients with undifferentiated paroxysmal spells. However, our results suggest that the specificity and sensitivity of FM as a marker for PNES is less than previously described among a large sample size of mixed general neurology patients. Acknowledgements: The content of this Abstract will be published in manuscript form in an upcoming volume of Epileptic Disorders. Reference: Benbadis SR. A spell in the epilepsy clinic and a history of "chronic pain" or "fibromyalgia" independently predict a diagnosis of psychogenic seizures. Epilepsy Behav 2005; 6(2): 264-5. Funding: No funding sources to report