Assessing Biases in the Diagnosis of Psychogenic Non-Epileptic Spells (PNES)
Ramsha Bhutta1, Ashley Bailey2, Marco Wolmarans3, Chindo Mallum4, Paul Grunsted3, Sarah Durica3, Denise Sebasigari3
1Department of Neurology, St. Francis-Emory Healthcare, 2Barrow Neurological Institute, 3University of Oklahoma Health Science Center, 4Department of Neurology, Jacobs School of Medicine and Biomedical Sciences
Objective:
Assessing types of biases in diagnosing psychogenic non-epileptic spells (PNES).
Background:
There is an established female predominance among patients diagnosed with PNES. While there have been neurobiological differences identified for this predominance, there lies the possibility of a misdiagnosis based on gender, increasing morbidity in these patients. Another survey found that African-Americans presenting with seizures are less likely to be admitted to the hospital or to undergo urgent neuroimaging compared to Caucasians, independent of insurance status.
Design/Methods:
A retrospective chart review was conducted at University of Oklahoma Health Sciences Center between January 2011 and December 2020 from electronic medical records (EMR). Patients greater than 5 years of age who were admitted to Epilepsy Monitoring Unit (EMU) were included and data regarding gender, race and education level was recorded using the REDCAP database system.
Results:
Of the 384 patients admitted to EMU, 62%(n=239) and 38 %(n=145) were women and men, respectively. 52%(n=122) of women were suspected to have PNES on initial visit out of which 50%(n=61) were concordant with both initial and EMU diagnosis of PNES. 36%(n=51) of men were suspected to have PNES in office out of which 43%(n=22) fell into the concordant group (agreement in initially suspected and final EMU diagnosis of PNES). 49%(n=66) Caucasians, 36%(n=5) African-Americans, 38%(n=8) patients with college level education and 27%(n=3) with elementary level education were in the concordant group. These differences were statistically significant in all of the aforementioned categories (McNemar’s Exact test, all p-values <0.002).
Conclusions:
There appears to be a small but significant difference in diagnosis of PNES depending on gender, age and race. Further large multi-center studies are needed to evaluate the differences. Once such biases and health disparities are uncovered, we can identify and address them ensuring accurate diagnosis leading to treatment and management in a timely manner reducing morbidity for our patients.