Assessing the personal impact of epilepsy in a population-based cohort of Veterans

Purpose: Epilepsy impacts patient lives in multidimensional ways. Although previous work has investigated epilepsy impact on health status, little is known about the overall quantified impact of epilepsy in Veterans. Our goal was to describe the impact of epilepsy on Veterans' lives using the Personal Impact of Epilepsy Scale (PIES) and determine the patient and clinical characteristics most strongly correlated with epilepsy impact. We described cohort characteristics and developed regression models to determine which characteristics were most strongly associated with PIES subscale (seizure, medication, comorbidity) scores and quality of life (QOL). Results: Approximately 36% of those who were invited responded to the survey. Of the 438 respondents included in the analyses, roughly 50% were aged 45–64 years (35% > 65; 14% 18–44); 19% were women. Almost 90% had previously received care by an epilepsy specialist, 37% of which was in Veterans Health Administration (VHA) and 38% in both VHA and community. The PIES scores were moderately low (mean: 88.68, [standard deviation (SD) = 63.24]; 300 total). The PIES overall and subscale scores were significantly lower for older Veterans with epilepsy (VWE) (> 65) compared with younger (18–44 years) and middle-aged (45–64 years) VWE [p < 0.001], indicating that older Veterans had a lower epilepsy impact overall, and for seizures, medication, and comorbidity. The younger and middle-aged VWE had a significantly higher proportion with psychiatric diagnosis compared with older VWE [p < 0.001]. There was a trend for significance for the overall PIES scores by gender, with women having total higher (worse) scores (mean = 93.10, SD = 69.68) than men (mean = 74.39, SD = 59.97), which was driven by a statistically higher score on the seizure subscale for women (mean = 27.66, SD = 27.97) compared with men (mean = 19.29, SD 25.35; p = 0.04). Regression models revealed that frequent seizures (> 1/month, > 2/month) and diagnoses of dementia significantly predicted higher (more negative) Seizure Severity PIES score [all p < 0.05]. Frequent seizures (> 1/month), number of antiepileptic drugs (AEDs), and diagnosis of dementia predicted negative impact, and older age predicted positive impact for medication subscale. Frequent seizures (> 1/month) and diagnoses of depression and dementia predicted negative mood and social impact [all p < 0.05]. Seizure frequency (> 2/month) was the only variable that significantly predicted lack of excellent QOL [p < 0.05]. Effects for gender were not significant after controlling for other variables. Conclusions: Findings were similar to a prior study of generic health outcomes in younger and older VWE using the 36-Item Short Form Survey (SF-36). Seizure frequency was consistently associated with negative impact of epilepsy in all age groups. While dementia and other diagnosed health conditions also contributed to epilepsy impact, older VWE had significantly lower PIES scores even after controlling for physical conditions and dementia. Lower (better) scores for comorbidity and medication scales in older VWE may be due to fewer diagnosed psychiatric comorbidities and psychiatric medication that have similar cognitive impact as AEDs, and which may also interact with AEDs. Implementation of patient self-management programs to improve seizure control may reduce epilepsy impact for Veterans and reduce Veterans Affairs (VA) healthcare utilization. The PIES may also be useful to measure outcomes of self-management interventions.