Rationale:
Psychogenic non-epileptic seizures (PNES) are commonly encountered in the clinic and hospital setting. The care for patients with PNES results in elevated healthcare costs and intense resource utilization. An electronic medical record (EMR) alarm was implemented at Rush University Medical Center in 2016 for all patients with the diagnosis of PNES to notify providers of the diagnosis and provide management guidelines while minimizing potential iatrogenic interventions.
A prior study showed the alarm is effective in preventing iatrogenic interventions and is safe to use in the hospital setting. This study aims to evaluate the effectiveness of the alarm in reducing healthcare utilization and redirect resources to benefit patients with PNES.
Methods:
A retrospective chart review was performed on all patients who presented to Rush University Medical Center between January 2017 and December 2019 who had the PNES alarm triggered. Data from number of outpatient clinic visits (neurology, psychology/psychiatry, and other specialties), diagnostic studies (Xray, CT, MRI, EKG, EEG, urine drug screen), and acute rehabilitation admissions were compared from the year before the first time the alarm was triggered for each individual patient with the year after the first alarm trigger. The time of the first alarm trigger was used as a surrogate for the time of the diagnosis made and documented in the EMR. Wilcoxon signed-rank test and McNemar test were used for statistical analysis. The current procedural terminology (CPT) codes were used for estimation of average costs for certain diagnostic studies and healthcare visits.
Results:
A total of 178 patients had the PNES alarm triggered during the study time. Mean age was 41, 80% were female, and 37% of patients also had a concurrent diagnosis of epilepsy. The median number of diagnostic studies done before first alarm trigger was 7 (0-40). Of the diagnostic studies done before first alarm trigger, 38% were CT, 27% Xray, 22% EKG, 8% EEG, 3% MRI and 2% urine drug screen. There was a statistically significant (p < 0.05) decrease in number of diagnostic studies, neurology clinic visits, and other non-neurologic/non-psychiatric clinic visits during the year after first alarm trigger. There was an increase in the number of psychiatry and psychology visits after the first alarm trigger but was not found to be statistically significant. One patient had an acute rehab admission before first alarm trigger but no patients after first alarm trigger.